Reverse desiliconization method for high-alumina fly ash

Abstract

The invention relates to the technical field of treating nonmetal raw materials to recover metal compounds, in particular to a reverse desiliconization method for high-alumina fly ash, which comprises the following steps: constructing a multi-stage countercurrent contact system, enabling the high-alumina fly ash to flow from a first-stage desiliconization tank to a final stage stage by stage, and enabling alkali liquor to reversely flow to form gradient distribution of final-stage low concentration and preceding-stage high concentration; the last-stage aluminum oxide dissolution is inhibited; the previous-stage silicon dioxide dissolution is enhanced. Performing multi-source data fusion analysis on the data to generate a slurry viscosity and liquid level incidence matrix, dynamically adjusting the rotating speed of a stirrer and the flow of an alkali liquid pump in combination with a PID algorithm, rehearsing the aluminum-silicon ratio and the solid content change trend by a virtual simulation model, and iteratively generating an optimized parameter combination of the alkali liquid flow rate, the underflow valve opening and the rotating speed of the stirrer through a genetic algorithm; according to the method, through the synergistic effect of physical separation and intelligent regulation and control, the secondary reaction intensity is reduced, and the aluminum-silicon separation efficiency and the process energy efficiency are improved.

Description

Technical Field

[0001] The invention relates to the technical field of processing non-metallic raw materials to recover metal compounds, in particular to a reverse desiliconization method for high-aluminum fly ash. Background Art

[0002] Existing high-alumina fly ash desiliconization processes, using reactors and casing equipment for desiliconization, result in high concentrations of silica and alumina in the slurry. Desiliconization is a vigorous secondary reaction, resulting in relatively low aluminum and silicon content in the desiliconized fly ash. When high-alumina fly ash is desiliconized once, the slurry is separated and washed, and then desiliconized a second time, the secondary desiliconization solution concentration is extremely low, resulting in a complex production process and high costs. Therefore, a novel reverse desiliconization process is needed to desiliconize high-alumina fly ash. This process can shorten the process, enable larger equipment, reduce equipment requirements, and lower production costs. Summary of the Invention

[0003] In view of the shortcomings of the existing technology, the present invention provides a reverse desiliconization method for high-aluminum fly ash. The present invention solves the problems of the existing high-aluminum fly ash desiliconization process, which uses a casing to heat the slurry for desiliconization, resulting in high concentrations of silica and alumina in the desiliconization slurry, making it difficult to inhibit the secondary reaction, low aluminum-silicon separation efficiency, poor filtration performance of the desiliconization product and low process energy efficiency.

[0004] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0005] The present invention provides a method for reverse desiliconization of high-aluminum fly ash, comprising:

[0006] Step 1: construct a multi-stage countercurrent contact system, wherein high-alumina fly ash is fed from the feed port of the first-stage desiliconization tank connected in series and flows step by step to the final-stage desiliconization tank, and alkali solution is injected from the top of the final-stage desiliconization tank and flows step by step in the reverse direction to the first-stage desiliconization tank, forming an alkali solution concentration gradient with low concentration in the final stage and high concentration in the previous stage;

[0007] Step 2: Real-time collection of slurry rheological property data, liquid level data, and pressure data is performed by using a viscosity sensor, liquid level gauge, and pressure transmitter installed in each stage of the desiliconization tank, and the real-time collection of slurry rheological property data, liquid level data, and pressure data is transmitted to a central controller via an industrial bus;

[0008] Step 3: adjusting the speed of the agitator in the desiliconization tank based on the slurry rheological property data, and controlling the flow output of the alkali liquid pump through a PID algorithm based on the liquid level data and pressure data to maintain the mass transfer stability of the reverse flow path;

[0009] Step 4: constructing a virtual simulation model based on the slurry rheological property data, liquid level data, and pressure data, simulating the changing trends of the aluminum-silicon ratio and solid content during the desiliconization process through the virtual simulation model, and generating an optimized parameter combination of alkali liquid flow rate, underflow valve opening, and agitator speed;

[0010] Step 5: reversely adjust the opening of the underflow valve of the preceding desiliconization tank according to the optimized parameter combination, and complete the coordinated desiliconization of the multi-stage sedimentation tank by controlling the material residence time.

[0011] Furthermore, in the reverse desiliconization method of high-alumina fly ash of the present invention, step 1 comprises:

[0012] The multi-stage countercurrent contact system controls the reverse flow rate through overflow weirs and underflow valves, wherein:

[0013] The height of the overflow weir is dynamically adjusted according to the viscosity value in the slurry rheological property data collected in step 2;

[0014] The opening of the underflow valve is controlled by a pneumatic regulating valve in conjunction with a PLC controller, and the PLC controller outputs a control instruction based on the liquid level data and pressure data collected in step 2;

[0015] The alkali solution concentration gradient distribution of low concentration in the final stage and high concentration in the preceding stage is used to suppress the dissolution of alumina in the final desiliconization tank and drive the virtual simulation model in step 4 to predict the aluminum-silicon ratio.

[0016] Furthermore, the reverse desiliconization method of high-alumina fly ash of the present invention, in step 2, comprises:

[0017] The slurry rheological property data is collected in real time by a viscosity sensor, the liquid level data is collected by a liquid level meter, and the pressure data is collected by a pressure transmitter;

[0018] The central controller receives the slurry rheological property data, liquid level data and pressure data through the industrial bus, and performs multi-source data fusion analysis to generate a slurry viscosity and liquid level correlation matrix;

[0019] The liquid level data is used to determine the overload risk of the overflow weir and trigger the PID algorithm control logic of step 3;

[0020] The slurry rheological property data and the liquid level data are used together as input parameters for adjusting the agitator speed in step 3.

[0021] Furthermore, the reverse desiliconization method of high-alumina fly ash of the present invention, in step 3, comprises:

[0022] The agitator adopts a double-layer blade design, the upper layer is an anchor blade for stabilizing the mass transfer interface of the reverse flow path in step 1, and the lower layer is a turbine blade for enhancing solid-liquid mixing;

[0023] When the viscosity value in the slurry rheological property data is higher than a preset threshold, a high shear blade mode is triggered to improve mixing efficiency;

[0024] The PID algorithm outputs a flow regulation instruction of the alkali liquid pump according to the liquid level data and pressure data in step 2;

[0025] The associated feedback of the pressure data and the liquid level data is used to correct the proportional coefficient and the integral time parameter of the PID algorithm.

[0026] Furthermore, the reverse desiliconization method of high-alumina fly ash of the present invention, in step 4, comprises:

[0027] The virtual simulation model integrates a fluid dynamics module and a chemical reaction dynamics module to simulate the change trend of the aluminum-silicon ratio and the solid content based on the slurry rheological property data, liquid level data and pressure data in step 2;

[0028] The simulation preview result of the aluminum-silicon ratio change trend is compared with the target aluminum-silicon ratio threshold value to generate an optimized parameter combination of alkali solution flow rate, underflow valve opening and agitator speed;

[0029] The optimized parameter combination is transmitted to the underflow valve regulating module in step 5 via the industrial bus.

[0030] Furthermore, the reverse desiliconization method of high-alumina fly ash of the present invention, in step 4, comprises:

[0031] The optimized parameter combination is iteratively generated by a genetic algorithm, wherein the genetic algorithm uses the aluminum-silicon ratio and the solid content as objective functions and performs parameter optimization based on the output results of the virtual simulation model in step 4;

[0032] The crossover rate and mutation rate of the genetic algorithm are dynamically adjusted according to the aluminum-silicon ratio deviation value in the simulation preview result;

[0033] When the simulation preview results deviate from the target aluminum-silicon ratio threshold, the parameter optimization process of the genetic algorithm is triggered;

[0034] The parameter adjustment instruction includes a reverse adjustment amount of the opening of the underflow valve of the preceding desiliconization tank, and the reverse adjustment amount is transmitted to the preceding desiliconization tank through the reverse flow path of step 1.

[0035] Furthermore, the reverse desiliconization method of high-alumina fly ash of the present invention, in step 5, comprises:

[0036] When the solid content of the final settling tank output by the virtual simulation model in step 4 is lower than the set threshold, the opening of the underflow valve of the preceding desiliconization tank is increased to shorten the material residence time;

[0037] When the solid content is higher than a set threshold, the opening of the bottom flow valve of the preceding desiliconization tank is reduced to extend the material residence time;

[0038] The optimized parameter combination is reversely transmitted to the underflow valve regulating module of the preceding desiliconization tank through the reverse flow path of step 1;

[0039] The collaborative desiliconization of the multi-stage sedimentation tank forms a closed-loop feedback based on the final-stage monitoring data of step 2 and the preceding control instructions of step 5.

[0040] Beneficial effects of the present invention:

[0041] The present invention significantly improves the efficiency of the high-aluminum fly ash desiliconization process through the synergistic effect of the reverse concentration gradient design of the multi-stage countercurrent contact system, data-driven dynamic regulation and closed-loop feedback mechanism. The high-concentration alkali solution in the reverse flow path preferentially dissolves the silica in the unreacted fly ash, and the low-concentration alkali solution in the final stage inhibits the dissolution of alumina, reducing the occurrence of secondary reactions through the physical separation reaction stage; the multi-source data fusion analysis of the viscosity sensor, liquid level gauge and pressure transmitter generates real-time feedback on the rheological characteristics and mass transfer state of the slurry, combined with the PID algorithm and the dynamic adjustment of the double-layer paddle agitator to optimize the mixing efficiency and maintain the mass transfer stability of the reverse flow; the virtual simulation model integrates fluid dynamics and chemical reaction kinetics modules, previews the changing trends of the aluminum-silicon ratio and solid content based on real-time data, generates a globally optimized parameter combination through genetic algorithm, and reversely regulates the opening of the bottom flow valve of the preceding desiliconization tank to match the material residence time. The above technical features synergistically reduce the excessive dissolution rate of alumina, improve the aluminum-silicon separation efficiency and product filtration performance, and at the same time reduce equipment redundancy and energy consumption through dynamic coordination between multi-stage tanks, solving the problems of violent secondary reactions, complex processes and low energy efficiency in existing processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, for ordinary technicians in this field, other drawings can be obtained based on the drawings without paying any creative labor.

[0043] Figure 1 A flow chart of a reverse desiliconization method for high-alumina fly ash provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. The technical solutions provided by each embodiment of the present invention are described in detail below in conjunction with the drawings. In order to better understand the purpose of the present invention, the present invention is further described in detail below.

[0045] See also Figure 1 The present invention provides a method for reverse desiliconization of high-aluminum fly ash, comprising:

[0046] Step 1: construct a multi-stage countercurrent contact system, wherein high-alumina fly ash is fed from the feed port of the first-stage desiliconization tank connected in series and flows step by step to the final-stage desiliconization tank, and alkali solution is injected from the top of the final-stage desiliconization tank and flows step by step in the reverse direction to the first-stage desiliconization tank, forming an alkali solution concentration gradient with low concentration in the final stage and high concentration in the previous stage;

[0047] In the reverse desiliconization method for high-aluminum fly ash described in the present invention, the construction of a multi-stage countercurrent contact system achieves dynamic regulation of the reverse concentration gradient through the coordination of physical structure and control logic. The system consists of a series of first-stage to last-stage desiliconization tanks. High-aluminum fly ash is continuously input from the feed port of the first-stage desiliconization tank and flows step by step to the last-stage desiliconization tank through an overflow weir. Alkali solution is injected from the top of the last-stage desiliconization tank and flows step by step to the first-stage desiliconization tank through an underflow valve along a reverse path. The reverse flow path adjusts the slurry flow rate through the linkage control of the overflow weir height and the underflow valve opening, wherein the overflow weir height is dynamically adjusted based on the viscosity value in the slurry rheological property data collected in step 2. When the viscosity increases, the overflow weir height is increased to slow the fly ash flow rate, and when the viscosity decreases, the height is reduced to accelerate the flow. The opening of the underflow valve is driven by a pneumatic control valve. The PLC controller generates control instructions based on the liquid level data and pressure data collected in real time in step 2. When the liquid level exceeds the preset upper limit, the opening is increased to reduce the pressure in the tank. When the liquid level is lower than the lower limit, the opening is reduced to maintain pressure stability.

[0048] The concentration gradient of the alkali solution with low concentration in the final stage and high concentration in the preceding stage is achieved through the physical design of the reverse flow path. The high-concentration alkali solution preferentially contacts the unreacted fly ash in the initial desiliconization stage, quickly dissolving silica to improve the desiliconization efficiency; the low-concentration alkali solution in the final stage reacts with the partially desiliconized material, reducing the dissolution rate of alumina and inhibiting the occurrence of secondary reactions. This gradient distribution is formed by the concentration difference between adjacent desiliconization tanks. The alkali concentration of the preceding tank body increases through the step-by-step accumulation of the reverse flow path, and the alkali concentration of the final tank body decreases step by step due to continuous reactions. The concentration gradient provides the input basis for the virtual simulation model of step 4, in which the solubility characteristics of the high-concentration alkali solution and the inhibitory effect of the low-concentration alkali solution are integrated into the chemical reaction kinetics module to predict the trend of the aluminum-silicon ratio change.

[0049] The mass transfer interface stability of the multi-stage countercurrent contact system is achieved through a double-layer control mechanism. The physical structural design of the overflow weir ensures the continuity of slurry flow and prevents uneven concentration caused by local retention. The dynamic adjustment of the underflow valve is based on real-time feedback from liquid level and pressure data to match the mass transfer requirements of different desiliconization stages. The synergistic effect of the concentration gradient and mass transfer stability of the reverse flow path physically blocks the local enrichment of high-concentration slurry, reduces the conditions for the occurrence of side reactions, and provides a predictable process environment for subsequent data collection and dynamic regulation.

[0050] Step 2: Real-time collection of slurry rheological property data, liquid level data, and pressure data is performed by using a viscosity sensor, liquid level gauge, and pressure transmitter installed in each stage of the desiliconization tank, and the real-time collection of slurry rheological property data, liquid level data, and pressure data is transmitted to a central controller via an industrial bus;

[0051] In the reverse desiliconization method of high-aluminum fly ash described in the present invention, step 2 realizes real-time monitoring and control of process parameters through a multi-source data acquisition and transmission mechanism. The viscosity sensor installed in each stage of the desiliconization tank is embedded in the inner wall of the tank body, directly contacting the slurry to collect rheological characteristic data in real time; the liquid level meter is fixed on the top of the tank body to monitor the slurry liquid level; the pressure transmitter is integrated into the bottom pipe of the tank to detect reaction pressure fluctuations. The sensor data is transmitted to the central controller in a periodic polling manner via an industrial bus (such as the Modbus RTU protocol). The CRC check mechanism is used to verify the integrity of the data during the transmission process. When the check fails, the retransmission process is triggered to avoid control deviations caused by communication anomalies.

[0052] The central controller performs multi-source data fusion analysis on the received slurry rheological property data, liquid level data and pressure data, extracts key characteristic parameters through principal component analysis, and generates a slurry viscosity-liquid level correlation matrix. The matrix quantifies the dynamic relationship between viscosity and liquid level, which is used to determine the overload risk of the overflow weir. When the liquid level data exceeds the preset safety range, the central controller generates an early warning signal and triggers the PID algorithm control logic of step 3, sending a flow adjustment instruction to the alkali liquid pump. After normalization, the slurry rheological property data and liquid level data are used as input parameters for adjusting the agitator speed. The corresponding relationship between the speed and the slurry viscosity is matched through the fuzzy control algorithm. For example, when the viscosity is high, the speed is increased to enhance the shear force.

[0053] The correlation feedback of the pressure data and liquid level data is achieved through time series modeling. When the pressure fluctuation exceeds the expected range, the central controller resets the integral term of the PID algorithm to avoid error accumulation and control instability. The viscosity value in the slurry rheological characteristic data is processed by sliding window mean filtering, and after eliminating instantaneous noise interference, it is input into the agitator speed adjustment module to ensure the balance between mixing efficiency and energy consumption. The data fusion result is transmitted back to the overflow weir height adjustment module in step 1 through the industrial bus, forming a closed-loop feedback across steps to support the dynamic stability of the multi-stage countercurrent system.

[0054] Step 3: adjusting the speed of the agitator in the desiliconization tank based on the slurry rheological property data, and controlling the flow output of the alkali liquid pump through a PID algorithm based on the liquid level data and pressure data to maintain the mass transfer stability of the reverse flow path;

[0055] In the reverse desiliconization method of high-aluminum fly ash described in the present invention, step 3 achieves mass transfer stability through the synergistic effect of dynamic control of the agitator and flow regulation of the alkali liquid pump. The agitator adopts a double-layer blade design, and the upper anchor blade matches the gap between the inner wall of the desiliconization tank. The mass transfer interface of the reverse flow path is stabilized by the circumferential shear force to prevent the slurry from stratifying; the lower turbine blade is located at the bottom of the tank body, and generates a radial flow field through high-speed rotation to enhance the solid-liquid mixing effect. When the viscosity value in the slurry rheological property data collected in step 2 exceeds the preset threshold, the central controller triggers the high shear blade mode, increases the turbine blade speed and switches the anchor blade rotation direction, breaks the agglomerated structure of the high-viscosity slurry by enhancing the shear force, and accelerates the dissolution of silica.

[0056] The PID algorithm generates an alkali liquid pump flow adjustment instruction based on the liquid level data and pressure data transmitted in step 2. The liquid level data is generated by real-time monitoring of the liquid level height in the desiliconization tank and a flow increase or decrease signal is generated in combination with a preset safety range; the pressure data reflects the reaction intensity in the tank. When the pressure increases, the alkali liquid injection volume is quickly reduced through the proportional term, and when the pressure decreases, the flow is gradually restored through the integral term. The output instruction of the PID algorithm is transmitted to the variable frequency drive of the alkali liquid pump via the industrial bus, and the motor speed is adjusted to achieve precise flow control. The correlation feedback of the pressure data and the liquid level data is realized through the time series analysis method. The central controller models the time series relationship between the liquid level fluctuation and the pressure change. When the liquid level rise rate does not match the pressure growth trend, the parameter correction logic of the PID algorithm is triggered to dynamically adjust the proportional coefficient and the integral time parameter.

[0057] The closed-loop feedback mechanism ensures process stability through real-time data acquisition in step 2 and dynamic control in step 3. The corrected PID parameters are then transmitted back to the agitator control module to optimize the coordinated action of the anchor and turbine blades. The enhanced mixing efficiency in high-shear mode, coupled with lye pump flow regulation, maintains uniform mass transfer in the reverse flow path, preventing secondary reactions caused by localized excess concentration.

[0058] Step 4: constructing a virtual simulation model based on the slurry rheological property data, liquid level data, and pressure data, simulating the changing trends of the aluminum-silicon ratio and solid content during the desiliconization process through the virtual simulation model, and generating an optimized parameter combination of alkali liquid flow rate, underflow valve opening, and agitator speed;

[0059] In the reverse desiliconization method of high-aluminum fly ash described in the present invention, step 4 realizes dynamic preview and control of the process through the construction of a virtual simulation model and parameter optimization. The virtual simulation model integrates a fluid dynamics module and a chemical reaction kinetics module, wherein the fluid dynamics module simulates the concentration gradient distribution and slurry flow rate field in the reverse flow path based on the slurry rheological characteristics data collected in step 2, and the chemical reaction kinetics module calculates the aluminum-silicon ratio change and solid content accumulation rate of the desiliconization reaction in combination with the liquid level data and pressure data. The two modules realize two-way data interaction through a coupling solver, and the flow rate distribution output by the fluid dynamics serves as the input boundary condition of the chemical reaction kinetics. The substance concentration changes generated by the chemical reaction kinetics are fed back to the fluid model to update the flow parameters, forming an iterative optimization simulation mechanism.

[0060] The simulation preview of the change trend of the aluminum-silicon ratio is achieved through a time-stepping algorithm, simulating the step-by-step evolution process of the aluminum-silicon ratio in a multi-stage desiliconization tank. The simulation results are compared with the preset target aluminum-silicon ratio threshold in real time, and the parameter optimization process is triggered when the preview value deviates from the threshold. The generation of the optimized parameter combination is based on a genetic algorithm, with the aluminum-silicon ratio and solid content as dual objective functions, and the parent generation individuals are screened by the roulette wheel selection method, and the offspring parameter combination is generated by simulated binary crossover and polynomial mutation operators. The crossover rate and mutation rate of the genetic algorithm are dynamically adjusted according to the aluminum-silicon ratio deviation value output by the simulation. When the deviation is large, the mutation rate is increased to enhance the global search capability, and when the deviation is small, the crossover rate is reduced to converge to the local optimal solution.

[0061] The generated optimized parameter combination includes the alkali liquid flow rate, the bottom flow valve opening and the agitator speed, which are transmitted to the bottom flow valve adjustment module in step 5 through the industrial bus. The OPC UA protocol is used during the transmission process to ensure the real-time and reliability of the data. The parameter instructions drive the pneumatic actuator to adjust the bottom flow valve opening after redundancy verification. The optimized parameters are transmitted back to the preceding desiliconization tank through the reverse flow path of step 1, and the material residence time is dynamically matched with the solid content threshold of the final sedimentation tank to form a closed-loop logic from simulation preview to actual control. This mechanism improves the aluminum-silicon separation efficiency and process stability through the synergistic effect of multi-stage tanks and dynamic adaptation of parameters.

[0062] Step 5: reversely adjust the opening of the underflow valve of the preceding desiliconization tank according to the optimized parameter combination, and complete the coordinated desiliconization of the multi-stage sedimentation tank by controlling the material residence time.

[0063] In the reverse desiliconization method of high-aluminum fly ash described in the present invention, step 5 realizes the coordinated desiliconization of multi-stage sedimentation tanks through the synergistic effect of closed-loop feedback of optimization parameters and the reverse flow path. The optimized parameter combination is generated by the virtual simulation model of step 4, including the alkali liquid flow rate, the bottom flow valve opening and the agitator speed, and is reversely transmitted to the bottom flow valve adjustment module of the preceding desiliconization tank through the industrial bus along the reverse flow path of step 1. During the transmission process, the central controller performs a redundant check on the parameter instructions. When the check fails, the historical parameter cache value is called to replace the abnormal data to avoid process fluctuations. The adjustment instruction of the bottom flow valve opening is realized by driving the valve plate displacement through a pneumatic actuator. The displacement of the actuator and the opening correction value are converted through a linear mapping relationship. The control accuracy reaches ±0.5%, ensuring the dynamic matching of the valve opening of the preceding tank body and the optimized parameters.

[0064] The control of material residence time is based on the solid content threshold judgment of the final settling tank. When the final solid content output in step 4 is lower than the set threshold, the central controller increases the opening of the bottom flow valve of the preceding desiliconization tank, shortens the material residence time to increase the desiliconization reaction rate; when the solid content is higher than the threshold, the opening is reduced to extend the residence time to inhibit excessive dissolution of alumina. The setting of the threshold is combined with historical process data and material characteristics, and is dynamically calibrated through the adaptive module of the central controller to adapt to the desiliconization requirements of different batches of fly ash. The adjustment logic derives the correction amount of the preceding valve through the concentration gradient relationship of adjacent tanks in the reverse flow path of step 1 to ensure that the parameter adjustment is consistent with the physical constraints of the reverse flow characteristics.

[0065] The coordinated desiliconization of the multi-stage sedimentation tank is achieved through closed-loop feedback of the final-stage monitoring data and the preceding control instructions. The solid content data of the final-stage sedimentation tank is collected in real time by the liquid level gauge and pressure transmitter in step 2, transmitted to the central controller via the industrial bus, and subjected to deviation analysis with the preview results of the virtual simulation model. When the deviation value exceeds the tolerance range, the parameter feedback mechanism is triggered, and the genetic algorithm optimization process of step 4 is restarted to generate an updated optimization parameter combination. The closed-loop feedback aligns the data timing through timestamp synchronization technology to ensure the dynamic consistency of the control instructions and the final-stage monitoring data. The control results are dynamically fed back to the simulation model through the agitator speed and alkali liquid pump flow in step 3, forming a complete control chain of "monitoring-simulation-execution", thereby improving the global stability and adaptability of the multi-stage desiliconization process.

[0066] The present invention provides a method for reverse desiliconization of high-aluminum fly ash, which achieves efficient desiliconization through the construction of a multi-stage countercurrent contact system and data-driven dynamic regulation. When constructing the multi-stage countercurrent contact system, high-aluminum fly ash is continuously input from the feed port of the first-stage desiliconization tank in series, and flows step by step to the final desiliconization tank through the overflow weir, while the alkali solution flows step by step from the top of the final desiliconization tank to the first-stage desiliconization tank in reverse, forming an alkali solution concentration gradient distribution with low concentration in the final stage and high concentration in the front stage. The design of the reverse flow path is achieved through the coordinated control of the overflow weir and the bottom flow valve, wherein the height of the overflow weir is dynamically adjusted according to the viscosity of the slurry, and the opening of the bottom flow valve is adjusted through the linkage of the pneumatic control valve and the PLC controller to maintain a stable mass transfer interface. This concentration gradient distribution suppresses the excessive dissolution of alumina in the final desiliconization tank and provides a physical basis for subsequent data acquisition and simulation optimization.

[0067] During the real-time data acquisition phase, viscosity sensors, level gauges, and pressure transmitters installed in each desiliconization tank simultaneously collect slurry rheological data, level data, and pressure data. This data is transmitted to a central controller via an industrial bus. Multi-source data fusion and analysis generate a slurry viscosity-level correlation matrix. Level data is used to determine the overload risk of the overflow weir in real time, triggering subsequent control logic. Pressure data is correlated with level data and fed back into the control algorithm. Both slurry rheological data and level data serve as input parameters for agitator speed adjustment, enabling the linkage of data acquisition and dynamic control.

[0068] During dynamic control, the agitator utilizes a dual-paddle design. The upper anchor-type blades stabilize the mass transfer interface in the reverse flow path, while the lower turbine-type blades enhance solid-liquid mixing. When the viscosity value in the slurry's rheological properties exceeds a preset threshold, the agitator switches to high-shear propeller mode to improve mixing efficiency. The caustic soda pump's flow output is controlled by a PID algorithm. This algorithm generates flow control commands based on liquid level and pressure data, and dynamically modifies the proportional coefficient and integral time parameters through feedback from the correlation between pressure and liquid level, ensuring a balance between mass transfer stability and energy efficiency.

[0069] The virtual simulation model is based on real-time data collected from the slurry's rheological properties, liquid level, and pressure, integrating fluid dynamics and chemical reaction kinetics modules. The model simulates the changing trends of the aluminum-silicon ratio and solids content during the desiliconization process. By comparing simulation pre-training results with the target aluminum-silicon ratio threshold, it generates optimized parameter combinations for caustic soda flow rate, underflow valve opening, and agitator speed. The optimized parameters are iteratively generated using a genetic algorithm, using the aluminum-silicon ratio and solids content as objective functions. The crossover rate and mutation rate are dynamically adjusted based on the deviation values in the simulation pre-training results, ensuring the global and adaptable nature of the parameter optimization process.

[0070] During the parameter execution phase, the optimized parameter combination is transmitted to the underflow valve adjustment module via the industrial bus. When the solid content of the final settling tank output by the virtual simulation model is lower than the set threshold, the opening of the underflow valve of the preceding desiliconization tank is increased to shorten the material residence time; when the solid content is higher than the threshold, the opening is reduced to extend the residence time. The parameters are transmitted back to the preceding desiliconization tank through the reverse flow path, and a closed-loop feedback loop is formed by combining the final-stage monitoring data with the preceding control instructions to achieve coordinated desiliconization of multiple-stage settling tanks. This closed-loop logic dynamically matches the simulation preview results with the actual operating conditions, improving the accuracy and stability of process control.

[0071] Specifically, the reverse desiliconization method of high-alumina fly ash of the present invention, step 1, comprises:

[0072] The multi-stage countercurrent contact system controls the reverse flow rate through overflow weirs and underflow valves, wherein:

[0073] The height of the overflow weir is dynamically adjusted according to the viscosity value in the slurry rheological property data collected in step 2;

[0074] The opening of the underflow valve is controlled by a pneumatic regulating valve in conjunction with a PLC controller, and the PLC controller outputs a control instruction based on the liquid level data and pressure data collected in step 2;

[0075] The alkali solution concentration gradient distribution of low concentration in the final stage and high concentration in the preceding stage is used to suppress the dissolution of alumina in the final desiliconization tank and drive the virtual simulation model in step 4 to predict the aluminum-silicon ratio.

[0076] In the reverse desiliconization method for high-alumina fly ash described herein, the construction of a multi-stage countercurrent contact system achieves precise regulation of the reverse flow rate through the coordinated control of an overflow weir and an underflow valve. The height of the overflow weir is dynamically adjusted based on the viscosity value of the slurry rheological properties data collected in real time by the viscosity sensor in step 2. When the slurry viscosity increases, the overflow weir height is increased to slow the slurry flow rate, and when the viscosity decreases, the overflow weir height is decreased to accelerate the flow. This adjustment mechanism, which receives viscosity data via a PLC controller and outputs adjustment commands to the overflow weir drive device, dynamically matches the slurry flow rate with the viscosity characteristics.

[0077] The underflow valve's opening is controlled by a pneumatic control valve linked to a PLC controller. The PLC generates control instructions based on the liquid level data collected by the level gauge and the pressure data collected by the pressure transmitter in step 2. When the liquid level exceeds a preset threshold, the PLC controller increases the underflow valve opening via the pneumatic control valve to reduce the pressure within the tank. When the liquid level falls below the threshold, the PLC controller decreases the opening to maintain pressure balance within the tank. The correlated feedback of pressure and liquid level data is used to modify the pneumatic control valve's operating parameters, ensuring real-time adaptation of the underflow valve opening to the slurry flow requirements.

[0078] The concentration gradient distribution of the alkali solution with low concentration in the final stage and high concentration in the preceding stage is achieved through the physical design of the reverse flow path. The high-concentration alkali solution preferentially contacts the unreacted fly ash in the initial desiliconization stage, quickly dissolving silica to improve the desiliconization efficiency; the low-concentration alkali solution in the final stage reacts with the partially desiliconized powder to reduce the dissolution rate of alumina, thereby inhibiting the intensity of the secondary reaction. This concentration gradient distribution also provides an input basis for the virtual simulation model of step 4, in which the dissolution characteristics of the high-concentration alkali solution in the preceding stage and the inhibitory effect of the low-concentration alkali solution in the final stage are integrated into the fluid dynamics module to predict the trend of the aluminum-silicon ratio change and optimize the output accuracy of the simulation model. Through the deep coupling of the reverse flow path and the data-driven model, dynamic coordination of process parameters and actual working conditions is achieved.

[0079] Specifically, the reverse desiliconization method of high-alumina fly ash of the present invention, in step 2, comprises:

[0080] The slurry rheological property data is collected in real time by a viscosity sensor, the liquid level data is collected by a liquid level meter, and the pressure data is collected by a pressure transmitter;

[0081] The central controller receives the slurry rheological property data, liquid level data and pressure data through the industrial bus, and performs multi-source data fusion analysis to generate a slurry viscosity and liquid level correlation matrix;

[0082] The liquid level data is used to determine the overload risk of the overflow weir and trigger the PID algorithm control logic of step 3;

[0083] The slurry rheological property data and the liquid level data are used together as input parameters for adjusting the agitator speed in step 3.

[0084] In the high-alumina fly ash reverse desiliconization method described herein, step 2 enables real-time monitoring and control of process parameters through multi-source data acquisition and fusion analysis. A viscosity sensor, level gauge, and pressure transmitter respectively collect real-time data on slurry rheological properties, liquid level, and pressure. The viscosity sensor is embedded in the inner wall of the desiliconization tank for direct contact with the slurry, the level gauge is installed at the top of the tank to monitor the liquid level, and the pressure transmitter is integrated into the bottom pipe of the tank to detect the reaction pressure. This data is transmitted to a central controller via an industrial bus, using the Modbus RTU standard for transmission protocol to ensure data synchronization and anti-interference capabilities.

[0085] The central controller performs multi-source data fusion analysis on the received slurry rheological data, liquid level data, and pressure data. Principal component analysis is used to extract key characteristic parameters and generate a slurry viscosity-liquid level correlation matrix. This matrix is used to quantify the dynamic relationship between viscosity and liquid level. When the liquid level data exceeds a preset safety range, an overflow weir overload warning signal is triggered, and a PID algorithm control instruction is sent to step 3 via the PLC controller. After normalization, the slurry rheological data and liquid level data serve as input parameters for agitator speed adjustment. A fuzzy control algorithm is used to match the corresponding relationship between speed and slurry viscosity.

[0086] The overload risk judgment logic of the liquid level data sets a dynamic threshold based on historical process data. When the liquid level data exceeds the upper threshold, the central controller generates an emergency speed reduction instruction and transmits it to the alkali liquid pump control module. The correlation analysis of pressure data and liquid level data is achieved through time series modeling. When the pressure fluctuation exceeds the expected range, the integral term of the PID algorithm is triggered to reset to avoid the accumulation of control errors. The viscosity value in the slurry rheological characteristic data is processed by sliding window mean filtering, and after eliminating instantaneous noise interference, it is input into the agitator speed adjustment module to achieve a balance between mixing efficiency and energy consumption. The data fusion result is transmitted back to the overflow weir height adjustment module in step 1 through the industrial bus, forming a closed-loop feedback across steps to support the dynamic stability of the multi-stage countercurrent system.

[0087] Specifically, the reverse desiliconization method of high-alumina fly ash of the present invention, in step 3, comprises:

[0088] The agitator adopts a double-layer blade design, the upper layer is an anchor blade for stabilizing the mass transfer interface of the reverse flow path in step 1, and the lower layer is a turbine blade for enhancing solid-liquid mixing;

[0089] When the viscosity value in the slurry rheological property data is higher than a preset threshold, a high shear blade mode is triggered to improve mixing efficiency;

[0090] The PID algorithm outputs a flow regulation instruction of the alkali liquid pump according to the liquid level data and pressure data in step 2;

[0091] The associated feedback of the pressure data and the liquid level data is used to correct the proportional coefficient and the integral time parameter of the PID algorithm.

[0092] In the reverse desiliconization method for high-alumina fly ash described in the present invention, step 3 improves mass transfer stability and mixing efficiency through agitator structural optimization and dynamic control algorithms. The agitator utilizes a double-layer blade design. The upper anchor blades match the clearance between the inner wall of the desiliconization tank, stabilizing the mass transfer interface in the reverse flow path through circumferential shearing to prevent slurry stratification. The lower turbine blades, located at the bottom of the tank, rotate at high speed to generate a radial flow field, enhancing the solid-liquid mixing effect. This structural design ensures that the slurry maintains a uniform concentration distribution during flow, providing a stable physical environment for the subsequent desiliconization reaction.

[0093] When the viscosity value in the slurry rheological properties data collected in step 2 exceeds a preset threshold, the central controller triggers the high-shear blade mode. In this mode, the speed of the turbine blade is increased to a preset upper limit, and the rotation direction of the anchor blade is switched to the reverse direction. By increasing the shear force, the agglomerated structure of the high-viscosity slurry is broken down and the dissolution of silica is accelerated. The viscosity threshold is set based on historical process data and the slurry rheological properties curve, and is dynamically calibrated by the central controller's adaptive module to adapt to the material property differences of different batches of fly ash.

[0094] The PID algorithm generates flow control instructions for the alkali liquid pump based on the liquid level and pressure data transmitted in step 2. The liquid level data is generated by real-time monitoring of the liquid level in the desiliconization tank, combined with a preset safety range, to generate flow rate increase or decrease signals. The pressure data reflects the intensity of the reaction within the tank. When the pressure rises, the proportional term rapidly reduces the alkali liquid injection rate, while when the pressure decreases, the integral term gradually restores the flow rate. The output instructions of the PID algorithm are transmitted via the industrial bus to the alkali liquid pump's variable frequency drive, which adjusts the motor speed to achieve precise flow control.

[0095] The correlation feedback between pressure data and liquid level data is achieved through time series analysis methods. The central controller models the time series relationship between liquid level fluctuations and pressure changes. When the liquid level rise rate does not match the pressure growth trend, the parameter correction logic of the PID algorithm is triggered. The proportional coefficient is dynamically adjusted according to the pressure gradient, and the integral time parameter is recalculated based on the cumulative error of the liquid level data. This correction mechanism avoids the control deviation caused by a single data source and improves the response speed and stability of the alkali solution flow regulation. The corrected PID parameters are fed back to the agitator control module through closed-loop feedback, forming a coordinated optimization of mixing efficiency and mass transfer stability.

[0096] Specifically, the reverse desiliconization method of high-alumina fly ash of the present invention, in step 4, comprises:

[0097] The virtual simulation model integrates a fluid dynamics module and a chemical reaction dynamics module to simulate the change trend of the aluminum-silicon ratio and the solid content based on the slurry rheological property data, liquid level data and pressure data in step 2;

[0098] The simulation preview result of the aluminum-silicon ratio change trend is compared with the target aluminum-silicon ratio threshold value to generate an optimized parameter combination of alkali solution flow rate, underflow valve opening and agitator speed;

[0099] The optimized parameter combination is transmitted to the underflow valve regulating module in step 5 via the industrial bus.

[0100] In the reverse desiliconization method of high-aluminum fly ash described in the present invention, step 4 realizes dynamic preview and precise control of process parameters through the construction and optimization of a virtual simulation model. The virtual simulation model integrates a fluid dynamics module and a chemical reaction kinetics module, wherein the fluid dynamics module simulates the concentration gradient distribution and slurry flow rate field in the reverse flow path based on the slurry rheological characteristics data collected in step 2, and the chemical reaction kinetics module calculates the aluminum-silicon ratio change and solid content accumulation rate of the desiliconization reaction in combination with the liquid level data and pressure data. The two modules realize data interaction through a coupling solver, wherein the flow rate distribution output by fluid dynamics serves as the input boundary condition of the chemical reaction kinetics, and the substance concentration change generated by the reaction kinetics is fed back to the fluid model to update the flow parameters, forming a bidirectional iterative simulation mechanism.

[0101] The simulation preview of the change trend of the aluminum-silicon ratio is achieved through a time-stepping algorithm to simulate the step-by-step evolution process of the aluminum-silicon ratio in a multi-stage desiliconization tank. The simulation results are compared with the preset target aluminum-silicon ratio threshold in real time. When the preview value deviates from the threshold, the parameter optimization process is triggered. The generation of the optimized parameter combination is based on the gradient descent method to search for the optimal solution, where the adjustment range of the alkali solution flow rate is limited by the physical constraints of the reverse flow path in step 1, and the correlation between the bottom flow valve opening and the agitator speed is determined by regression analysis modeling of historical process data. The parameter combination is transmitted to the bottom flow valve adjustment module in step 5 via the industrial bus. The transmission protocol adopts the OPC UA standard to support reliable interaction between real-time data synchronization and remote control instructions.

[0102] The execution logic of the optimized parameter combination is deeply coupled with the reverse flow design of step 1. The adjustment instruction of the alkali liquid flow rate controls the motor speed of the alkali liquid pump through the frequency converter, the adjustment amount of the bottom flow valve opening is achieved by driving the valve plate displacement through the pneumatic actuator, and the optimized value of the agitator speed is converted into a pulse signal output by the servo controller. During the parameter transmission process, the central controller performs redundancy check on the data packets of the industrial bus. When the check fails, the historical parameter rollback mechanism is triggered to avoid process fluctuations caused by communication anomalies. The dynamic application of the optimized parameters forms a closed loop through the feedback loop of step 5 and the real-time data acquisition of step 2, realizing dynamic calibration of simulation preview and actual working conditions, and improving the control accuracy and stability of the desiliconization process.

[0103] Specifically, the reverse desiliconization method of high-alumina fly ash of the present invention, in step 4, comprises:

[0104] The optimized parameter combination is iteratively generated by a genetic algorithm, wherein the genetic algorithm uses the aluminum-silicon ratio and the solid content as objective functions and performs parameter optimization based on the output results of the virtual simulation model in step 4;

[0105] The crossover rate and mutation rate of the genetic algorithm are dynamically adjusted according to the aluminum-silicon ratio deviation value in the simulation preview result;

[0106] When the simulation preview results deviate from the target aluminum-silicon ratio threshold, the parameter optimization process of the genetic algorithm is triggered;

[0107] The parameter adjustment instruction includes a reverse adjustment amount of the opening of the underflow valve of the preceding desiliconization tank, and the reverse adjustment amount is transmitted to the preceding desiliconization tank through the reverse flow path of step 1.

[0108] In the reverse desiliconization method for high-aluminum fly ash described in the present invention, the parameter optimization process in step 4 achieves global optimization of process parameters through deep interaction between a genetic algorithm and a virtual simulation model. The genetic algorithm uses the aluminum-silicon ratio and solid content as objective functions and constructs a fitness function through a multi-objective weighted strategy, wherein the weight coefficient of the aluminum-silicon ratio is dynamically allocated based on the alkali solution concentration gradient in the reverse flow path of step 1. The output results of the virtual simulation model serve as the input data set for the genetic algorithm, including the aluminum-silicon ratio distribution of the multi-stage desiliconization tank, the accumulated solid content, and the mass transfer efficiency parameters of the reverse flow path. After data normalization to eliminate dimensional differences, they are input into the algorithm iteration process.

[0109] The genetic algorithm's crossover and mutation rates are dynamically adjusted based on the aluminum-silicon ratio deviation from simulation pre-run results. When the aluminum-silicon ratio deviation exceeds a preset threshold, the crossover rate is gradually reduced based on the convergence trend of historical iteration results, reducing random perturbations in the solution space. The mutation rate is positively correlated with the deviation through an exponential function, enhancing local search capabilities and helping to escape local optima. This dynamic adjustment mechanism is implemented by the central controller's adaptive module, which has an embedded deviation-parameter mapping table to match the optimal combination of crossover and mutation rates in real time.

[0110] When the simulation preview results deviate from the target aluminum-silicon ratio threshold, the central controller triggers the genetic algorithm's parameter optimization process. This triggering condition is based on sliding window statistics, generating an optimization instruction when the average aluminum-silicon ratio output from three consecutive simulations exceeds the threshold. Once the parameter optimization process is initiated, the genetic algorithm selects parent individuals using a roulette wheel selection method, generates offspring parameter combinations using a simulated binary crossover operator, and introduces randomness through a polynomial mutation operator to iteratively generate optimized parameter sets for caustic soda flow rate, underflow valve opening, and agitator speed.

[0111] The reverse adjustment amount of the bottom flow valve opening of the preceding desiliconization tank in the parameter adjustment instruction is reversely transmitted through the reverse flow path constructed in step 1. The calculation of the adjustment amount is based on the optimal parameter combination output by the genetic algorithm, and the valve opening correction value of the preceding tank body is derived through the concentration gradient relationship of the adjacent desiliconization tanks in the reverse flow path. The correction value is transmitted to the pneumatic actuator of the preceding desiliconization tank via the industrial bus to drive the precise adjustment of the bottom flow valve opening. During the parameter transmission process, the central controller performs boundary verification on the adjustment amount, and automatically cuts it off to a reasonable range when it exceeds the physical constraints to avoid equipment overload. The reverse adjustment mechanism works synergistically with the closed-loop feedback of step 5 to form a complete process chain from final-stage monitoring to preceding-stage regulation, thereby realizing dynamic balance and efficient coordination of multi-stage desiliconization tanks.

[0112] Specifically, the reverse desiliconization method of high-alumina fly ash of the present invention, in step 5, comprises:

[0113] When the solid content of the final settling tank output by the virtual simulation model in step 4 is lower than the set threshold, the opening of the underflow valve of the preceding desiliconization tank is increased to shorten the material residence time;

[0114] When the solid content is higher than a set threshold, the opening of the bottom flow valve of the preceding desiliconization tank is reduced to extend the material residence time;

[0115] The optimized parameter combination is reversely transmitted to the underflow valve regulating module of the preceding desiliconization tank through the reverse flow path of step 1;

[0116] The collaborative desiliconization of the multi-stage sedimentation tank forms a closed-loop feedback based on the final-stage monitoring data of step 2 and the preceding control instructions of step 5.

[0117] In the reverse desiliconization method of high-aluminum fly ash described in the present invention, step 5 forms a closed-loop feedback through the monitoring data of the final sedimentation tank and the control instructions of the preceding desiliconization tank, thereby realizing dynamic coordination of multiple-stage sedimentation tanks. When the solid content of the final sedimentation tank output by the virtual simulation model of step 4 is lower than the set threshold, the central controller generates an instruction to increase the opening of the bottom flow valve of the preceding desiliconization tank, shortening the residence time of the material in the tank and increasing the desiliconization reaction rate; when the solid content is higher than the threshold, the opening is reduced to extend the residence time and inhibit excessive dissolution of alumina. The setting of the threshold is based on historical process data and material characteristics, and is dynamically calibrated by the adaptive module of the central controller to adapt to the desiliconization requirements under different working conditions.

[0118] The optimized parameter combination is reversely transmitted to the bottom flow valve adjustment module of the preceding desiliconization tank through the reverse flow path constructed in step 1. During the parameter transmission process, the industrial bus uses a redundant check mechanism to verify the integrity of the data packet. When the check fails, the cache value of the preceding parameter is called to replace the abnormal data to avoid process fluctuations. The adjustment instruction of the bottom flow valve opening is realized by driving the valve plate displacement by a pneumatic actuator. The displacement of the pneumatic actuator and the opening correction value are converted through a linear mapping relationship, and the control accuracy reaches ±0.5%. The physical design of the reverse transmission path is based on the series structure of the multi-stage desiliconization tank. The valve opening correction value of the preceding tank body is derived through the concentration gradient correlation between adjacent tank bodies to ensure that the parameter adjustment matches the reverse flow characteristics.

[0119] The coordinated desiliconization of the multi-stage sedimentation tank forms a closed-loop feedback through the final monitoring data of step 2 and the preceding control instructions of step 5. The solid content data of the final sedimentation tank is collected in real time by the pressure transmitter and liquid level gauge of step 2, transmitted to the central controller via the industrial bus, and subjected to deviation analysis with the preview results of the virtual simulation model. When the deviation value exceeds the tolerance range, the parameter feedback mechanism is triggered, and the genetic algorithm optimization process of step 4 is restarted to generate an updated optimization parameter combination. The closed-loop feedback mechanism realizes data alignment through timestamp synchronization technology to ensure the temporal consistency of the preceding control instructions and the final monitoring data. The execution results of the control instructions are dynamically fed back to the virtual simulation model through the agitator speed and alkali liquid pump flow of step 3, forming a complete control chain of "monitoring-simulation-optimization-execution", thereby improving the stability and adaptability of the multi-stage desiliconization process.

[0120] A multi-stage countercurrent contact system consists of multiple desiliconization tanks connected in series. High-alumina fly ash is continuously fed into the feed port of the first-stage desiliconization tank and flows step by step to the final desiliconization tank. Alkali liquor flows in the reverse direction from the top of the final desiliconization tank to the first-stage desiliconization tank. This system uses the reverse flow path to create an alkali liquor concentration gradient, with low concentration in the final stage and high concentration in the initial stage. This difference in concentration preferentially dissolves silica, inhibits alumina dissolution, and reduces secondary reactions.

[0121] Reverse flow path: This design involves the opposite flow of high-aluminum fly ash and alkali solution in the desiliconization tank. Fly ash flows sequentially from the first stage to the final stage, while alkali solution flows in the opposite direction. The flow rate is regulated by the coordinated control of an overflow weir and underflow valve. This path physically separates the reaction stages, avoiding side reactions caused by localized overconcentration, and improving aluminum-silicon separation efficiency.

[0122] Slurry rheological properties data: Slurry flow characteristic parameters, including viscosity and shear rate, collected in real time by the viscosity sensor. This data is used to determine the slurry mixing state, trigger agitator speed adjustments (such as switching to high shear mode for high viscosity), and optimize solid-liquid mixing uniformity.

[0123] Virtual simulation model: A numerical model integrating fluid dynamics (CFD) and chemical reaction kinetics simulates the changing trends of the aluminum-silicon ratio and solid content during the desiliconization process based on real-time data collected from the slurry's rheological properties, liquid level, and pressure. The model output is used to predict process outcomes and generate optimized parameter combinations (such as caustic soda flow rate and underflow valve opening) using a genetic algorithm.

[0124] Genetic Algorithm: A heuristic optimization algorithm that uses the aluminum-silicon ratio and solid content as objective functions and iteratively generates the optimal parameter combination by simulating the biological evolution process (selection, crossover, and mutation). The crossover and mutation rates are dynamically adjusted based on simulation deviations to avoid falling into local optimal solutions.

[0125] Closed-loop feedback mechanism: This mechanism dynamically coordinates the solids content monitoring data from the final settling tank with the control instructions from the preceding desiliconization tank. When the solids content deviates from the threshold, the opening of the preceding tank's underflow valve is adjusted in the opposite direction (for example, increasing the opening to shorten the residence time). The optimized parameters are then transmitted to the execution module via the industrial bus, enabling dynamic calibration of process parameters.

[0126] PID algorithm: A proportional-integral-differential control algorithm outputs flow control instructions for the caustic soda pump based on liquid level and pressure data. It reduces flow when the liquid level exceeds the limit and adjusts the proportional coefficient and integral time parameters when the pressure fluctuates to maintain reverse flow mass transfer stability.

[0127] Slurry Viscosity and Level Correlation Matrix: A two-dimensional data model generated by a central controller through multi-source data fusion quantifies the dynamic relationship between viscosity and level. This matrix is used to determine the risk of weir overload and serves as an input for agitator speed adjustments.

[0128] Double-layer impeller design: The agitator's structural design features upper anchor-type impellers that stabilize the mass transfer interface and prevent slurry stratification. Lower turbine-type impellers rotate at high speed to create a radial flow field, enhancing solid-liquid mixing. High viscosity conditions result in a high-shear mode, improving mixing efficiency.

[0129] Reverse flow path transfer: This involves transferring optimized parameters (such as underflow valve opening) back to the preceding desiliconization tank via an industrial bus along the reverse flow path. For example, when the final stage solids holdup falls below a threshold, the preceding stage valve opening is increased to shorten material residence time, enabling coordinated control of multiple tanks.

[0130] The specific embodiment of the present invention solves the problems of severe secondary reaction, low separation efficiency and insufficient energy efficiency in the existing high-aluminum fly ash desiliconization process through the synergistic effect of a multi-stage countercurrent contact system, data-driven control and closed-loop feedback mechanism. During implementation, the first to last stage desiliconization tanks connected in series constitute a multi-stage countercurrent contact system. High-aluminum fly ash is continuously input from the first stage feed port and flows step by step to the last stage through the overflow weir; the alkali solution flows back from the top of the last stage desiliconization tank to the first stage, forming a gradient distribution of low concentration in the last stage (for example, 0.5-1.2 mol / L) and high concentration in the previous stage (for example, 2.5-3.5 mol / L). This reverse flow path is achieved by dynamic adjustment of the overflow weir height and the bottom flow valve opening, wherein the overflow weir height is adjusted in real time according to the slurry viscosity value, and the bottom flow valve opening is controlled by a PLC controller based on the liquid level and pressure data, thereby suppressing the dissolution of alumina in the last stage desiliconization tank and improving the silica dissolution efficiency of the previous stage tank body.

[0131] The real-time data acquisition and dynamic control module is implemented through the viscosity sensor, liquid level gauge, and pressure transmitter embedded in the desiliconization tank. The viscosity sensor monitors the rheological properties of the slurry, the liquid level gauge collects the liquid level, and the pressure transmitter detects the reaction pressure in the tank. The data is transmitted to the central controller via the Modbus RTU protocol. The controller uses multi-source data fusion technology to construct a correlation matrix between the slurry viscosity and liquid level, triggering the PID algorithm to generate flow control instructions for the alkali liquid pump. The agitator adopts a double-layer blade design. The upper anchor blade maintains the stability of the mass transfer interface, and the lower turbine blade switches to high shear mode when the slurry viscosity exceeds the preset threshold (for example, 1500mPa·s), thereby improving mixing efficiency. The correlation feedback of the liquid level data and pressure data corrects the proportional coefficient and integral time parameters of the PID algorithm through time series modeling to ensure the stability of the mass transfer process.

[0132] The virtual simulation model integrates fluid dynamics and chemical reaction kinetics modules, simulating the changing trends of the aluminum-silicon ratio and solid content based on real-time data. It then iteratively generates optimized parameter combinations through a genetic algorithm. The genetic algorithm uses the aluminum-silicon ratio (target threshold ≥ 5) and solid content (target threshold ≥ 30%) as dual-objective functions. The crossover rate and mutation rate are dynamically adjusted based on the deviation of the aluminum-silicon ratio output by the simulation. Optimized parameters include the caustic soda flow rate (e.g., 0.8-1.5 m³ / h), underflow valve opening (e.g., 30-70%), and agitator speed (e.g., 200-400 rpm). These parameters are transmitted back to the underflow valve control module of the preceding desiliconization tank via an industrial bus. When the solid content of the final settling tank falls below the set threshold, the opening of the preceding valve is increased to shorten the residence time; when it exceeds the threshold, the opening is decreased to extend the reaction time. A closed-loop feedback mechanism dynamically matches the final-stage monitoring data with the preceding control instructions to achieve coordinated desiliconization of the multi-stage tanks, reducing energy consumption and improving aluminum-silicon separation efficiency and filtration performance.

[0133] The present invention solves the defects of the existing high-aluminum fly ash desiliconization process through a multi-stage countercurrent contact system and a data-driven dynamic control mechanism. First, a multi-stage countercurrent contact system is constructed, and the high-aluminum fly ash flows step by step from the first-stage desiliconization tank to the last stage, and the alkali solution flows in reverse to form a gradient distribution with low concentration in the last stage and high concentration in the front stage. The high-concentration alkali solution preferentially dissolves the silicon dioxide in the unreacted fly ash, thereby improving the desiliconization efficiency; the low-concentration alkali solution in the last stage inhibits the excessive dissolution of alumina and reduces the intensity of the secondary reaction. The reverse flow path dynamically adjusts the slurry flow rate and the stability of the mass transfer interface through the coordinated control of the overflow weir and the bottom flow valve, blocks the local accumulation of high-concentration slurry from a physical design perspective, and reduces the conditions for the occurrence of secondary reactions.

[0134] Secondly, real-time control based on multi-source data fusion optimizes aluminum-silicon separation efficiency and filtration performance. Viscosity sensors, level gauges, and pressure transmitters collect real-time data on slurry rheological properties, liquid level, and pressure. The central controller dynamically adjusts the agitator speed and alkali pump flow rate using a PID algorithm. The agitator utilizes a double-layered impeller design. Anchor-type impellers stabilize the mass transfer interface, while turbine-type impellers enhance solid-liquid mixing. High-viscosity slurries trigger a high-shear mode to improve mixing uniformity. Correlated feedback from liquid level and pressure data modifies the PID algorithm's control parameters, maintaining reverse flow mass transfer stability and thereby improving the particle distribution and filtration performance of the desiliconization product.

[0135] Finally, the virtual simulation model and the closed-loop feedback mechanism work together to improve process energy efficiency. The simulation model integrates fluid dynamics and chemical reaction kinetics, previews the changing trends of the aluminum-silicon ratio and solid content based on real-time data, and generates optimized parameter combinations for alkali liquid flow rate, underflow valve opening, and agitator speed through genetic algorithms. The optimized parameters are transmitted back to the preceding desiliconization tank through the reverse flow path, and the underflow valve opening is dynamically adjusted according to the solid content threshold of the final sedimentation tank, forming a closed-loop logic of "final-stage monitoring-precursor regulation". This mechanism reduces energy consumption and improves the global stability of process control through the synergistic effect of multi-stage tanks and dynamic matching of parameters, achieving the dual goals of efficient aluminum-silicon separation and energy efficiency optimization.

Claims

1. A method for reverse desiliconization of high-alumina fly ash, characterized in that: include: Step 1: construct a multi-stage countercurrent contact system, wherein high-alumina fly ash is fed from the feed port of the first-stage desiliconization tank connected in series and flows step by step to the final-stage desiliconization tank, and alkali solution is injected from the top of the final-stage desiliconization tank and flows step by step in the reverse direction to the first-stage desiliconization tank, forming an alkali solution concentration gradient with low concentration in the final stage and high concentration in the previous stage; Step 2: Real-time collection of slurry rheological property data, liquid level data, and pressure data is performed by using a viscosity sensor, liquid level gauge, and pressure transmitter installed in each stage of the desiliconization tank, and the real-time collection of slurry rheological property data, liquid level data, and pressure data is transmitted to a central controller via an industrial bus; Step 3: adjusting the speed of the agitator in the desiliconization tank based on the slurry rheological property data, and controlling the flow output of the alkali liquid pump through a PID algorithm based on the liquid level data and pressure data to maintain the mass transfer stability of the reverse flow path; Step 4: constructing a virtual simulation model based on the slurry rheological property data, liquid level data, and pressure data, simulating the changing trends of the aluminum-silicon ratio and solid content during the desiliconization process through the virtual simulation model, and generating an optimized parameter combination of alkali liquid flow rate, underflow valve opening, and agitator speed; Step 5: reversely adjust the opening of the underflow valve of the preceding desiliconization tank according to the optimized parameter combination, and complete the coordinated desiliconization of the multi-stage sedimentation tank by controlling the material residence time.

2. The reverse desiliconization method of high-alumina fly ash according to claim 1, characterized in that: The step 1 comprises: The multi-stage countercurrent contact system controls the reverse flow rate through overflow weirs and underflow valves, wherein: The height of the overflow weir is dynamically adjusted according to the viscosity value in the slurry rheological property data collected in step 2; The opening of the underflow valve is controlled by a pneumatic regulating valve in conjunction with a PLC controller, and the PLC controller outputs a control instruction based on the liquid level data and pressure data collected in step 2; The alkali solution concentration gradient distribution of low concentration in the final stage and high concentration in the preceding stage is used to suppress the dissolution of alumina in the final desiliconization tank and drive the virtual simulation model in step 4 to predict the aluminum-silicon ratio.

3. The reverse desiliconization method of high-alumina fly ash according to claim 1, characterized in that: The step 2 includes: The slurry rheological property data is collected in real time by a viscosity sensor, the liquid level data is collected by a liquid level meter, and the pressure data is collected by a pressure transmitter; The central controller receives the slurry rheological property data, liquid level data and pressure data through the industrial bus, and performs multi-source data fusion analysis to generate a slurry viscosity and liquid level correlation matrix; The liquid level data is used to determine the overload risk of the overflow weir and trigger the PID algorithm control logic of step 3; The slurry rheological property data and the liquid level data are used together as input parameters for adjusting the agitator speed in step 3.

4. The reverse desiliconization method of high-alumina fly ash according to claim 1, characterized in that: The step 3 includes: The agitator adopts a double-layer blade design, the upper layer is an anchor blade for stabilizing the mass transfer interface of the reverse flow path in step 1, and the lower layer is a turbine blade for enhancing solid-liquid mixing; When the viscosity value in the slurry rheological property data is higher than a preset threshold, a high shear blade mode is triggered to improve mixing efficiency; The PID algorithm outputs a flow regulation instruction of the alkali liquid pump according to the liquid level data and pressure data in step 2; The associated feedback of the pressure data and the liquid level data is used to correct the proportional coefficient and the integral time parameter of the PID algorithm.

5. The reverse desiliconization method of high-alumina fly ash according to claim 1, characterized in that: The step 4 includes: The virtual simulation model integrates a fluid dynamics module and a chemical reaction dynamics module to simulate the change trend of the aluminum-silicon ratio and the solid content based on the slurry rheological property data, liquid level data and pressure data in step 2; The simulation preview result of the aluminum-silicon ratio change trend is compared with the target aluminum-silicon ratio threshold value to generate an optimized parameter combination of alkali solution flow rate, underflow valve opening and agitator speed; The optimized parameter combination is transmitted to the underflow valve regulating module in step 5 via the industrial bus.

6. The reverse desiliconization method of high-alumina fly ash according to claim 5, characterized in that: The step 4 includes: The optimized parameter combination is iteratively generated by a genetic algorithm, wherein the genetic algorithm uses the aluminum-silicon ratio and the solid content as objective functions and performs parameter optimization based on the output results of the virtual simulation model in step 4; The crossover rate and mutation rate of the genetic algorithm are dynamically adjusted according to the aluminum-silicon ratio deviation value in the simulation preview result; When the simulation preview results deviate from the target aluminum-silicon ratio threshold, the parameter optimization process of the genetic algorithm is triggered; The parameter adjustment instruction includes a reverse adjustment amount of the opening of the underflow valve of the preceding desiliconization tank, and the reverse adjustment amount is transmitted to the preceding desiliconization tank through the reverse flow path of step 1.

7. The reverse desiliconization method of high-alumina fly ash according to claim 1, characterized in that: The step 5 includes: When the solid content of the final settling tank output by the virtual simulation model in step 4 is lower than the set threshold, the opening of the underflow valve of the preceding desiliconization tank is increased to shorten the material residence time; When the solid content is higher than a set threshold, the opening of the bottom flow valve of the preceding desiliconization tank is reduced to extend the material residence time; The optimized parameter combination is reversely transmitted to the underflow valve regulating module of the preceding desiliconization tank through the reverse flow path of step 1; The coordinated desiliconization of the multi-stage sedimentation tank is based on the final monitoring data of step 2 and the preceding control instructions of step 5.

Images