CN-122018450-A - Lime milk impurity removal system and method for calcium silicate production

CN122018450ACN 122018450 ACN122018450 ACN 122018450ACN-122018450-A

Abstract

The invention relates to the technical field of industrial automation and chemical engineering data processing, in particular to a lime milk impurity removal system and method for calcium silicate production, which are used for acquiring data in real time through a data acquisition module to generate a time-sequence multidimensional feature vector; the impurity dynamic deduction module simulates an impurity motion track based on a three-dimensional geometric model and outputs prediction data, the multi-target collaborative optimization decision module generates a pareto optimal process scheme by taking vibration frequency of a vibrating screen, stay time of a sedimentation tank, jump-sieve pore diameter combination and flocculant addition amount as decision variables, the dynamic scheduling module converts the scheme into executable instructions to control physical equipment to operate, the virtual-real interaction closed-loop correction module generates deviation signals by comparing actual measurement data and prediction data of an optical detection unit and a spectrum analyzer, and model parameters and weight coefficients are updated by utilizing a federal learning mechanism to form closed-loop correction.

Inventors

WU ENMIN
GAO PEIJUN
MENG FANYANG
LI LINFENG
HAO RUI
CHAI WEI
WANG HUIXING
WANG ZHIFEI
WU ZHUANG

Assignees

鄂尔多斯市蒙泰铝业有限责任公司

Dates

Publication Date: 20260512
Application Date: 20251225

Claims (10)

1. The lime milk impurity removing system for calcium silicate production is characterized by comprising a physical equipment device and a control device, wherein the control device is in communication connection with the physical equipment device; The physical equipment device comprises a lime ash reactor, a vibrating screen, a primary sedimentation tank, a jump screen, a two-stage sedimentation tank, a precision filter, a high-speed centrifuge and a chemical flocculation reactor which are connected in sequence, and an optical detection unit and a spectrum analyzer which are arranged at the outlet of the two-stage sedimentation tank; The control device comprises a data acquisition module, an impurity dynamic deduction module, a multi-target collaborative optimization decision-making module, a dynamic scheduling module and a virtual-real interaction closed loop correction module; The data acquisition module is configured to acquire data of particle size distribution, spectral characteristics, turbidity and pH value of lime milk from the physical equipment device, and perform denoising and time stamp alignment processing on the acquired data to generate a time-sequence multidimensional feature vector; The impurity dynamic deduction module is configured to receive the time-sequence multidimensional feature vector, simulate an impurity motion track through a computational fluid dynamics and discrete element method based on a three-dimensional geometric model of the physical equipment device, dynamically correct simulation parameters by utilizing a long-period memory network and output predicted data of impurity types, concentrations and evolution trend; the multi-target collaborative optimization decision-making module is configured to receive the prediction data, take the vibrating frequency of the vibrating screen, the stay time of the sedimentation tank, the combination of the jump sieve pore diameters and the addition amount of the flocculating agent as decision variables, and adopt a multi-target optimization algorithm to generate a pareto optimal process scheme; The dynamic scheduling module is configured to convert the pareto optimal process scheme into executable instructions and control each device in the physical device to operate; the virtual-real interaction closed-loop correction module is configured to compare measured data and predicted data of the optical detection unit and the spectrum analyzer to generate a deviation signal, update model parameters of the impurity dynamic deduction module through a federal learning mechanism, and adjust weight coefficients of the multi-target collaborative optimization decision module to form closed-loop correction.
2. The lime milk impurity removal system for calcium silicate production of claim 1, further comprising: The device comprises a data acquisition module, a time-sequence multidimensional feature vector input impurity dynamic deduction module, a multi-target collaborative optimization decision module, a pareto optimal process scheme input dynamic scheduling module, a physical equipment device driven by a control instruction of the dynamic scheduling module, an optical detection unit and actual measurement data of a spectrum analyzer, a virtual-real interaction closed-loop correction module, a deviation signal reverse adjustment impurity dynamic deduction module of the virtual-real interaction closed-loop correction module and the multi-target collaborative optimization decision module, and a collaborative network for data bidirectional circulation, wherein the prediction data of the impurity dynamic deduction module is input into the multi-target collaborative optimization decision module; the data acquisition module comprises an optical sensing unit, a chemical sensing unit and a data preprocessing unit; the optical sensing unit is configured to collect particle size distribution and spectral feature data through the laser particle size analyzer and the hyperspectral imaging probe; the chemical sensing unit is configured to collect turbidity and pH data via an online turbidity sensor and a pH electrode; the data preprocessing unit is configured to perform wavelet transformation denoising and sliding window time stamp alignment on the data acquired by the optical sensing unit and the chemical sensing unit, and generate a time-sequence multidimensional feature vector; The data of the optical sensing unit and the chemical sensing unit are input into the data preprocessing unit in parallel, and the output of the data preprocessing unit is used as the input of the impurity dynamic deduction module.
3. The lime milk impurity removal system for calcium silicate production of claim 1, wherein the impurity dynamic deduction module comprises a model construction layer, a simulation calculation layer and a learning optimization layer; The model construction layer is configured to establish a three-dimensional geometric model of the physical equipment device through the three-dimensional scanning data; The simulation calculation layer is configured to simulate lime milk flow and impurity movement through coupling of computational fluid mechanics and discrete element methods; the learning optimization layer is configured to dynamically correct simulation parameters by utilizing the long-term and short-term memory network to fuse historical production data; The three-dimensional geometric model of the model construction layer provides grid boundary conditions for the simulation calculation layer, a simulation result of the simulation calculation layer is input into the learning optimization layer, and output of the learning optimization layer is used as prediction data.
4. The lime milk impurity removal system for calcium silicate production of claim 1, wherein the multi-objective collaborative optimization decision-making module comprises a parameter coding mechanism, a fitness evaluation flow and a pareto solution screening strategy; the parameter encoding mechanism is configured to normalize the decision variables to a hybrid encoding chromosome; the fitness evaluation flow is configured to take impurity removal efficiency, energy consumption cost and medicament consumption as fitness functions; the pareto solution screening strategy is configured to select an optimal scheme through non-dominant sorting and congestion degree calculation; The parameter coding mechanism outputs and inputs the fitness evaluation flow, and the result of the fitness evaluation flow generates a pareto optimal process scheme through a pareto solution screening strategy.
5. The lime milk impurity removal system for calcium silicate production of claim 1, wherein the dynamic scheduling module comprises an instruction conversion unit and a real-time scheduling unit; the instruction conversion unit is configured to convert the pareto optimal process scheme into a PLC executable instruction; the real-time scheduling unit is configured to monitor the running state of the equipment through a reinforcement learning algorithm and adjust the parameters of the equipment in a self-adaptive manner; The output of the instruction conversion unit drives the physical equipment device, and the real-time scheduling unit collects equipment state data and feeds the equipment state data back to the instruction conversion unit to form a local closed loop.
6. The lime milk impurity removal system for calcium silicate production of claim 1, wherein the virtual-real interaction closed loop correction module comprises a deviation calculation mechanism and a federal learning process; the deviation calculating mechanism is configured to calculate the mahalanobis distance between the measured data and the predicted data through principal component analysis; The federal learning process is configured to encrypt the deviation data and aggregate updated model parameters; the deviation signal of the deviation calculation mechanism triggers the federal learning process, and the output of the federal learning process simultaneously updates the impurity dynamic deduction module and the multi-target collaborative optimization decision module.
7. The lime milk impurity removal system for calcium silicate production according to claim 1, wherein after the prediction data of the impurity dynamic deduction module is input into the multi-objective collaborative optimization decision-making module, the multi-objective collaborative optimization decision-making module calls the digital twin model interface simulation parameter effect, and the simulation result is fed back to the impurity dynamic deduction module to update the initial boundary condition, so as to form a virtual trial-and-error closed loop.
8. The lime milk impurity removal system for calcium silicate production according to claim 1, wherein the equipment operation state data of the dynamic scheduling module is synchronized to the impurity dynamic deduction module in real time, the flow field parameters are corrected, and meanwhile, the deviation signals of the virtual-real interaction closed loop correction module adjust the weight coefficients of the multi-objective collaborative optimization decision module to enable the system to adapt to raw material fluctuation.
9. The lime milk impurity removal system for calcium silicate production according to claim 1, wherein the sampling frequency of the data acquisition module is dynamically adjusted by the virtual-real interaction closed-loop correction module according to the deviation, the standby detection equipment is switched when sensor data are abnormal, and model parameters of the impurity dynamic deduction module and the multi-objective collaborative optimization decision module are shared across production lines through a federal learning mechanism for improving system robustness.
10. A lime milk impurity removal method for calcium silicate production, applied to the lime milk impurity removal system for calcium silicate production according to any one of claims 1 to 9, characterized by comprising: Step 1, collecting particle size distribution, spectral characteristics, turbidity and pH value data of lime milk from a physical equipment device, denoising and time stamp alignment processing are carried out on the collected data to generate a time-sequence multi-dimensional characteristic vector, wherein the physical equipment device comprises a lime ash reactor, a vibrating screen, a primary sedimentation tank, a jump screen, a two-stage sedimentation tank, a precise filter, a high-speed centrifuge and a chemical flocculation reactor which are sequentially connected, and an optical detection unit and a spectrum analyzer which are arranged at an outlet of the two-stage sedimentation tank; Step 2, receiving a time-sequence multidimensional feature vector, simulating an impurity motion track based on a three-dimensional geometric model of a physical equipment device by a computational fluid dynamics and discrete element method, dynamically correcting simulation parameters by utilizing a long-period memory network, and outputting predicted data of impurity types, concentrations and evolution trends; Step 3, receiving prediction data, taking the vibration frequency of a vibrating screen, the stay time of a sedimentation tank, the combination of the jump-sieve pore diameters and the addition amount of a flocculating agent as decision variables, and adopting a multi-objective optimization algorithm to generate a pareto optimal process scheme; Step 4, converting the pareto optimal process scheme into executable instructions, and controlling each device in the physical device to operate; step 5, comparing the measured data and the predicted data of the optical detection unit and the spectrum analyzer to generate a deviation signal, updating the simulation parameters in the step 2 through a federal learning mechanism, and simultaneously adjusting the weight coefficient in the step 3 to form closed loop correction; The time-sequence multidimensional feature vector generated in the step 1 is input into the step 2, the prediction data output in the step 2 is input into the step 3, the pareto optimal process scheme generated in the step 3 is input into the step 4, the control instruction of the step 4 drives a physical device, the actual measurement data of the optical detection unit and the spectrum analyzer are fed back to the step 5, the deviation signal generated in the step 5 reversely adjusts the parameters of the step 2 and the step 3, and a data bidirectional circulation cooperative network is formed for removing impurities in lime milk so as to improve the whiteness and purity of calcium silicate products.

Description

Lime milk impurity removal system and method for calcium silicate production Technical Field The invention relates to the technical field of industrial automation and chemical engineering data processing, in particular to a lime milk impurity removal system and method for calcium silicate production. Background In the existing calcium silicate production, quality control starts from high-purity screening of raw materials, the production of byproducts is reduced due to low impurity content, chemical uniformity of products is improved, temperature and pressure parameters in a reaction stage need to be accurately regulated and controlled to promote ordered formation of calcium silicate crystals, so that mechanical properties of the materials are optimized, an online monitoring technology tracks process variables in real time, operation conditions are timely adjusted through data feedback, stability of a production process is maintained, and a final product is subjected to multidimensional detection, including physical and chemical analysis, whether the final product meets application standards is verified, and reliable management of overall quality is achieved. The existing calcium silicate production quality control technology has the technical problems that the efficiency of the purification process of raw materials and process fluid is insufficient, for example, in the lime milk preparation process, raw burnt limestone and insufficiently burnt carbon particles are difficult to completely separate only through conventional sedimentation and simple filtration, fine particles of the raw burnt limestone and insufficiently burnt carbon particles can enter a synthesis process along with lime milk, meanwhile, dust impurities contained in recycled production sewage are subjected to sedimentation treatment, micron-sized suspended matters easily penetrate an existing filtration unit and are circularly enriched in a system, and colloid suspended matters in sodium silicate solution also lack an on-line monitoring and efficient removing means. The impurities are used as black spot causes and can be directly doped into a calcium silicate crystal structure to cause fluctuation of whiteness and purity indexes of a finished product, and the conventional detection method relying on off-line sampling has hysteresis and cannot perform real-time early warning and feedback control on impurity introduction, so that the quality stability of the product is difficult to maintain. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a lime milk impurity removal system and a lime milk impurity removal method for calcium silicate production, which solve the technical problems that the raw materials (lime milk and sodium silicate solution) for synthesizing calcium silicate and solid impurities and suspended matters existing in production sewage are difficult to effectively monitor and remove, so that black point defects appear in a final product, and the quality control index (such as whiteness and purity) of the product cannot be stably judged by a conventional detection method. In order to solve the technical problems, the invention comprises the following specific contents: In a first aspect, the invention provides a lime milk impurity removal system for calcium silicate production, which comprises a physical equipment device and a control device, wherein the control device is in communication connection with the physical equipment device; The physical equipment device comprises a lime ash reactor, a vibrating screen, a primary sedimentation tank, a jump screen, a two-stage sedimentation tank, a precision filter, a high-speed centrifuge and a chemical flocculation reactor which are connected in sequence, and an optical detection unit and a spectrum analyzer which are arranged at the outlet of the two-stage sedimentation tank; The control device comprises a data acquisition module, an impurity dynamic deduction module, a multi-target collaborative optimization decision-making module, a dynamic scheduling module and a virtual-real interaction closed loop correction module; The data acquisition module is configured to acquire data of particle size distribution, spectral characteristics, turbidity and pH value of lime milk from the physical equipment device, and perform denoising and time stamp alignment processing on the acquired data to generate a time-sequence multidimensional feature vector; The impurity dynamic deduction module is configured to receive the time-sequence multidimensional feature vector, simulate an impurity motion track through a computational fluid dynamics and discrete element method based on a three-dimensional geometric model of the physical equipment device, dynamically correct simulation parameters by utilizing a long-period memory network and output predicted data of impurity types, concentrations and evolution trend; the multi-target collaborative optimization decision-making