Intelligent optimization method for ore blending section of sulfuric acid method titanium dioxide production

Abstract

The application discloses a method for intelligently optimizing ore blending in sulfuric acid method titanium dioxide production, comprising the following steps: S1, an intelligent ore blending prediction model is established; the chemical component mass content of each component in each single ilmenite raw material is detected, and all detection data is accurately entered into an Excel table to establish a raw material component database; S2, an acidolysis reaction heat guidance model is established; inorganic oxide standard molar reaction enthalpy basic data is sorted to establish a parameter database, and a titanium ore acidolysis reaction heat calculation model is established in combination with raw material blending; S3, an economic model of intelligent ore blending is established; a linear programming model is constructed with the lowest cost per ton of titanium as an optimization target; S4, multi-dimensional constraint conditions are set; S5, an Excel built-in nonlinear programming solving tool is called to finally output an optimal ore blending scheme meeting all process constraint conditions and a corresponding lowest production cost per ton of titanium. The application has the beneficial effect that optimization is performed with the lowest cost per ton of titanium as a target, and the optimal cost is realized within all constraint boundaries.

Description

Technical Field

[0001] This invention relates to the field of sulfuric acid process titanium dioxide production technology, and in particular to a multi-mineral synergistic blending process for ilmenite, which is used to calculate the optimal ratio of multiple minerals and achieve optimal control of production costs in the blending process. Background Technology

[0002] Currently, the ore blending process for sulfuric acid titanium dioxide still relies on manual experience, which presents problems such as complex ore composition, numerous constraint indicators, high coupling degree of optimization objectives, and lagging adjustment of ore blending schemes. It is difficult to achieve low-cost, stable, and intelligent ore blending under the constraints of multiple indicators such as TiO2, Fe2O3, FeO, V2O5, P2O5, Nb2O5, MnO, and Cr2O3.

[0003] Chinese invention patent authorization announcement number CN113283707B discloses a method for determining the mixing ratio of high-phosphorus titanium ore and ordinary titanium ore. The method involves mixing high-phosphorus titanium ore and ordinary titanium ore to form a mixed ore. The formula for calculating the mixing ratio of high-phosphorus titanium ore and ordinary titanium ore is as follows: Mixing ratio = Upper limit of P index / 50% * m%. However, this method only constrains one impurity (P) and does not provide guidance for mixing ore under multiple index constraints. Summary of the Invention

[0004] To address the shortcomings of existing ore diversification, complex constraints, and low optimization efficiency, this invention provides an intelligent optimization method for the ore blending stage in sulfuric acid process titanium dioxide production. This method achieves multi-constraint, low-cost, and rapid ore blending of ilmenite to provide the optimal blending ratio for the acidolysis reaction. The method of this application effectively solves the limitations and uncertainties of current ore blending processes, such as slow and inaccurate processes and reliance on manual experience.

[0005] The technical solution of this invention is: an intelligent optimization method for the ore blending section in the sulfuric acid process for titanium dioxide production, comprising the following steps: S1, Establishment of an intelligent ore blending prediction model: Detecting the chemical composition and mass content of TiO2, Fe2O3, FeO, SiO2, Al2O3, CaO, MgO, S, P2O5, Nb2O5, MnO, V2O5, and Cr2O3 in each individual ilmenite raw material, and accurately entering all detection data into an Excel spreadsheet to establish a raw material component database; S2, Establishment of an acidolysis reaction heat-guided model: Organizing basic data on the standard molar enthalpy of inorganic oxides to establish a parameter database, and establishing a calculation model for the acidolysis reaction heat of ilmenite in conjunction with the raw material ratio; S3, Establishment of an intelligent ore blending economic model: With the minimum cost per ton of titanium as the optimization objective, constructing the following linear programming model: min C = Σ(a i ×C i In the formula: C: the comprehensive cost of titanium product per unit after ore blending, in yuan / ton of titanium; a i : The proportion of the i-th type of ilmenite (0≤xi≤1); C i: Unit titanium production cost when using only the i-th type of ilmenite raw material; S4, Multi-dimensional constraint setting: Delineate all constraint boundaries by combining production process indicators, impurity control, reaction conditions, and material usage ratios: Main grade constraint: Σ(a i ×TiO 2， i)≥44%~46%, impurity upper limit constraint: Σ(a i ×V2O 5， i)≤0.2%~0.3%, Σ(a i ×P2O 5， i)≤0.1%~0.15%, Σ(a i ×MnO ， i)≤0.5%~1.5%, Σ(a i ×Cr2O 3， i)≤0.1%~0.15%, Σ(a i ×Nb2O 5， i) ≤0.01%~0.05%, thermal constraint for acidolysis reaction: 800 MJ / t~880 MJ / t≤Q 混 ≤900 MJ / t~1000 MJ / t; Specific mineral ratio limits: a7≤0.05-0.1, a8≤0.05-0.1, a9≤0.3-0.5; S5, Call the built-in nonlinear programming solver in Excel, use the established economic model as the objective function, set various indicators as constraints, input the relevant solver parameters and start the calculation, and finally output the optimal mineral blending scheme that meets all process constraints and the corresponding lowest production cost per ton of titanium.

[0006] In the above scheme, S1 uses X-ray fluorescence spectroscopy to detect the mass content of chemical components TiO2, Fe2O3, FeO, SiO2, Al2O3, CaO, MgO, S, P2O5, Nb2O5, MnO, V2O5, and Cr2O3 in each type of ilmenite raw material.

[0007] The calculation model for the mass fraction of each chemical component in the mixed ore established in S1 of the above scheme is as follows: W TiO2 =Σ(a i ×w TiO2 ,i),W Fe2O3 =Σ(a i ×w Fe2O3， i), W FeO =Σ(a i ×w FeO， i), W Al2O3 =Σ(a i ×w Al2O3， i), W CaO =Σ(a i ×w CaO，i), W S =Σ(a i ×w S， i), W P2O5 =Σ(a i ×w P2O5， i), W Nb2O5 =Σ(a i ×w Nb2O5， i), W MnO =Σa i ×w MnO， i), W V2O5 =Σ(a i ×w V2O5， i), W Cr2O3 =Σ(a i ×w Cr2O3， i); where: W represents the mass fraction of the corresponding component in the mixture; a i w*,i represents the mass ratio of the i-th ilmenite raw material in the blend; w*,i represents the original mass fraction of the corresponding component in the i-th ilmenite raw material; and i represents the raw material ore type number.

[0008] The guiding model for the heat of reaction of ilmenite in S2 described in the above scheme is as follows: Heat of reaction of single ore Q i =∑(w k,i ÷M k ×Δr H m,k ), the combined reaction heat of the mixed ore Q 混 =Σ(x i ×Q i In the formula: Q 混 The heat of reaction per unit mass of acid hydrolysis of mixed ilmenite, MJ / t; Q i w represents the heat of acidolysis per unit mass of the i-th type of single ore; k,i Content of type k oxide component in type i ore; ΔrH m,k M represents the standard molar enthalpy of reaction of the corresponding oxide; k This represents the relative molecular mass of the corresponding oxide.

[0009] The beneficial effects of this invention are that it relies on the nonlinear programming solution function of Excel to input the calculation formulas of the three major models of component prediction, reaction heat calculation, and economic cost, along with the measured basic data, into the spreadsheet system. The built-in nonlinear programming operation module is then called to perform numerical solutions. Based on fully satisfying various process constraints, the optimal ore blending scheme is selected to determine the lowest comprehensive production cost. The optimization is carried out with the goal of minimizing the cost per ton of titanium, achieving cost optimization within all constraint boundaries. The calculation response speed is fast, with the ore blending scheme generated in no more than 2 minutes, which significantly improves efficiency compared to traditional manual ore blending. It can be adapted to various actual production scenarios, such as full availability of mineral resources, partial mineral resource restrictions, and mineral resource supply interruption replacement, and has strong stability and applicability. Detailed Implementation

[0010] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the examples. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments implemented by those skilled in the art without creative effort are within the protection scope of the present invention.

[0011] The present invention will be further described in detail below with reference to specific embodiments, so as to more clearly illustrate the technical solution and implementation effect of the present invention.

[0012] Example 1: This example uses the nonlinear programming solution function of the Excel platform to solve for the optimal ore blending scheme in the ore blending section of sulfuric acid process titanium dioxide production. The specific implementation steps are as follows: S1. Use X-ray fluorescence spectrometry (XRF) to detect the composition of each batch of ilmenite raw materials, accurately obtain the mass fraction data of TiO2, Fe2O3, FeO, SiO2, Al2O3, CaO, MgO, S, P2O5, Nb2O5, MnO, V2O5, and Cr2O3 in each type of ilmenite, and accurately enter all the detection data into an Excel spreadsheet to establish a raw material composition database; S2. In the Excel spreadsheet, according to the ore blending scheme constructed above... S3. Based on the acid hydrolysis reaction heat guidance model established earlier, combine the basic parameters such as the standard molar enthalpy of reaction and relative molecular mass of each oxide with the raw material component data entered in step 1, and enter the reaction heat calculation expression in Excel to complete the construction of the reaction heat model; S4. With the goal of minimizing the comprehensive production cost per ton of titanium, enter the economic optimization model expression constructed earlier in Excel to complete the construction of the cost model; S5. According to the actual production process requirements, set various constraints in Excel. The specific constraint values ​​are as follows: Main grade constraint: Σ(a i ×TiO 2， i)≥44%~46%, impurity upper limit constraint: Σ(a i ×V2O 5， i)≤0.2%~0.3%, Σ(a i ×P2O 5， i)≤0.1%~0.15%, Σ(a i ×MnO ， i)≤0.5%~1.5%, Σ(a i ×Cr2O 3， i)≤0.1%~0.15%, Σ(a i ×Nb2O 5，i) ≤0.01%~0.05%, thermal constraint for acidolysis reaction: 800 MJ / t~880 MJ / t≤Q 混 ≤900 MJ / t~1000 MJ / t; Specific mineral ratio limits: a7≤0.05-0.1, a8≤0.05-0.1, a9≤0.3-0.5; S6. Use Excel's built-in nonlinear programming solver, with the established economic model as the objective function (finding the minimum value), and the set indicators as constraints. After inputting the relevant solver parameters, start the calculation and finally output the optimal ore blending scheme that satisfies all process constraints and the corresponding lowest production cost per ton of titanium.

[0013] Example 2: Nine types of ilmenite were selected as blending raw materials. The content of each component, measured heat of reaction, and market unit price of each raw material were detected by X-ray fluorescence spectrometry (XRF). The basic data in Table 1 were compiled (unit: except for heat of reaction in MJ / t and unit price of ore in yuan / ton, all other components are mass fraction %):

[0014] The ore blending requirements are: all ore sources are available, satisfying the following: TiO2 ≥ 45%; V2O5 ≤ 0.28%; P2O5 ≤ 0.14%; Nb2O5 ≤ 0.05%; MnO ≤ 1.4%; Cr2O3 ≤ 0.15%; 850 ≤ Σ(x i • Q i a7≤0.07; a8≤0.07; a9≤0.4 and the lowest cost.

[0015] The intelligent ore blending model outputs the following results: A: 28%, B: 16%, C: 0%, D: 0%, E: 0%, F: 2%, G: 7%, H: 7%, I: 40%. A represents the first ore source, B represents the second ore source, and so on, with I representing the ninth ore source. The comprehensive blending unit price is 1770 yuan / ton. Index verification: TiO2: 45.17%×28%+46.27%×16%+46.52%×2%+42.18%×7%+45.13%×7%+45.67%×40%=45.36% (≥45%, meets the standard) V2O5: 0.16%×28%+0.2%×16%+0.15%×2%+0.2%×7%+0.17%×7%+0.13%×40%=0.18% (≤0.28%, meets the standard) P2O5: 0.26%×28%+0.04%×16%+0.06%×2%+0.18%×7%+0.09%×7%+0.09%×40%=0.135% (≤0.14%, meets the standard) Nb2O5: 0.03%×28%+0.08%×16%+0.04%×2%+0×7%+0%×7%+0.02%×40%=0.03% (≤0.05%, meets the standard) MnO: 0.96%×28%+1.27%×16%+0.58%×2%+0.86%×7%+0.94%×7%+0.97%×40%=1% (≤1.4%, meets the standard) Cr2O3: 0.08%×28%+0.05%×16%+0.01%×2%+0.01%×7%+0.01%×7%+0.01%×40%=0.04% (≤0.15%, meets the standard) Heat of reaction: 914.78×28%+888.56×16%+954.89×2%+967.24×7%+982.42×7%+1000.14×40%=953.71MJ / t (850-1000 MJ / t, meets the standard).

[0016] Under the premise that all mineral resources are available and all preset process constraints are met, this embodiment obtains the cost-optimized mineral blending scheme through intelligent mineral blending model, verifying the effectiveness, accuracy and practicality of the model of the present invention.

[0017] Example 3: This example simulates actual production conditions with limited mineral resources. Only three types of ilmenite (No. 1, No. 2, and No. 6) are selected for blending. Based on the established intelligent blending optimization model, the optimal cost ratio scheme is solved within the limited process indicators. Specific data are shown in Table 2.

[0018]

[0019] This ore blending is limited to the above three ore sources, and the following production process control constraints are set: TiO2 ≥ 45%; V2O5 ≤ 0.22%; P2O5 ≤ 0.135%; Nb2O5 ≤ 0.035%; MnO ≤ 1%; Cr2O3 ≤ 0.11%; 880 ≤ Q 混矿 With a yield of ≤980 MJ / t, the lowest overall production cost of ore blending is achieved while meeting all constraints.

[0020] The intelligent ore blending model outputs the following results: A: 49%, B: 19%, C: 32%, where A represents the first ore source, B represents the second ore source, and C represents the third ore source. The comprehensive blending unit price is 1912 yuan / ton. Index verification: TiO2: 45.17%×49%+46.27%×19%+47.09%×32%=45.99% (≥45%, meets the standard) V2O5: 0.16%×49%+0.2%×19%+0.15%×32%=0.164% (≤0.22%, meets the standard) P2O5: 0.26%×49%+0.04%×19%+0×32%=0.135% (≤0.135%, meets the standard) Nb2O5: 0.03%×49%+0.08%×19%+0.02×32%=0.04% (≤0.035%, meets the standard) MnO: 0.96%×49%+1.27%×19%+0.48%×32%=0.87% (≤1%, meets the standard) Cr2O3: 0.08%×49%+0.05%×19%+0.01%×32%=0.05% (≤0.1%, meets the standard) Heat of reaction: 914.78×49%+888.56×19%+998.73×32%=936.36MJ / t (880-980 MJ / t, meets the standard) This embodiment successfully adapts to production conditions where mineral resources are scarce and raw material selection is limited. While strictly adhering to various quality indicators for titanium dioxide production and the thermal conditions of the acidolysis process, it quickly derives feasible and economically optimal ore blending combinations, further verifying that the intelligent ore blending model of this invention has strong adaptability and high practicality, and can effectively cope with various abnormal raw material supply scenarios on site.

[0021] Comparative Example 4: This example serves as a counterexample to Example 3. It again selects three types of ilmenite (No. 1, No. 2, and No. 6) as raw materials for blending. Through three sets of typical manual / traditional experience-based blending ratios, the superiority of the intelligent blending model of this invention is verified. The three sets of ratios correspond to common problems in on-site blending such as excessive impurities, process failure, and non-compliance and uneconomical practices. Specific experimental and analytical results are as follows.

[0022] 1) Impurities exceed standards, product quality fails to meet standards. The following results will be output based on the following ore blending ratio: A: 60%, B: 10%, C: 30%, where A represents No. 1 ilmenite, B represents No. 2 ilmenite, and C represents No. 6 ilmenite. The comprehensive unit price of the blended ore is 1730×60%+2017×10%+2130×30%=1878.7 yuan / ton. The indicator verification is as follows: TiO2: 45.17%×60%+46.27%×10%+47.09%×30%=45.86% (≥45%, meets the standard); P2O5: 0.26%×60%+0.04%×10%+0×30%=0.160% (≥0.135%, not up to standard); The V2O5, Nb2O5, MnO, Cr2O3, and reaction heat parameters all meet the constraints.

[0023] This formulation is based on traditional experience and focuses solely on titanium content and reaction heat, neglecting stringent phosphorus constraints. Excessive P2O5 leads to excessively high phosphorus impurity content in the titanium dioxide product, severely interfering with subsequent calcination processes, affecting the calcination strength of the titanium dioxide particles, and causing quality defects such as reduced whiteness and a yellowish hue. This is a completely unproductive formulation. This counterexample demonstrates that manual, experience-based ore blending easily overlooks critical constraints on trace impurities and cannot balance multiple indicators simultaneously.

[0024] 2) If the component indicators are qualified but the process conditions are ineffective, the following results will be output according to the following ore blending ratio: A: 10%, B: 20%, C: 80%, where A represents No. 1 ilmenite, B represents No. 2 ilmenite, and C represents No. 6 ilmenite. The heat of reaction for mixing the ores is: 914.8 × 10% + 888.6 × 10% + 998.7 × 80% = 989.38 MJ / t (>980 MJ / t, exceeding the standard). All impurity indicators of TiO2 fully meet the process constraints. While all component indicators of this formulation meet the standards, the heat of reaction for mixing the ores is too high, exceeding the upper limit of thermal control for the acidolysis process. Excessive heat of reaction will lead to violent acidolysis, material splashing, and uncontrolled reaction temperature, posing a safety hazard of "explosion" and causing uneven decomposition of titanium ore and a reduced acidolysis rate. This results in the process being completely unable to operate stably, making it a typical ineffective formulation with acceptable components but an infeasible process.

[0025] 3) The process is compliant and feasible, but the production cost is not optimal. The following ore blending ratio produces the following results: A: 30%, B: 20%, C: 50%, where A represents No. 1 ilmenite, B represents No. 2 ilmenite, and C represents No. 6 ilmenite. All component indicators and heat of reaction meet the process constraints, and the blending ratio is compliant and can be put into normal production. The comprehensive unit price of the blended ore is: 1730×20%+2017×30%+2130×50%=2046.1 yuan / ton.

[0026] The manually selected compliant blending ratio, compared to the optimal solution output by the intelligent ore blending model of this invention (unit price 1912 yuan / ton), increases the production cost per ton by 134.4 yuan. Although this ratio meets both quality and process requirements, it does not maximize economic benefits, fully exposing the pain point of manual ore blending under conditions of limited ore resources and multiple constraints, which easily leads to maintaining quality while exceeding costs, resulting in unnecessary losses in enterprise production efficiency.

[0027] The three counterexamples in Comparative Example 4 cover three major categories of common on-site ore blending problems: excessive impurities, thermal failure of the process, and compliance without economy. They precisely address the shortcomings of manual ore blending and traditional experience-based ore blending. The results of the comparison with the optimal embodiment fully demonstrate that the intelligent ore blending model of this invention can uniquely output the optimal blending ratio that achieves "full compliance of indicators, feasible process, and lowest cost" under complex working conditions with limited ore resources and multiple coupled constraints. The model's adaptability, accuracy, and engineering practicality are significantly better than traditional ore blending methods.

Claims

1. A smart optimization method for the ore blending section in the sulfuric acid process titanium dioxide production, characterized by: Includes the following steps: S1. Establishment of Intelligent Ore Blending Prediction Model: The chemical composition and mass content of TiO2, Fe2O3, FeO, SiO2, Al2O3, CaO, MgO, S, P2O5, Nb2O5, MnO, V2O5, and Cr2O3 in each individual ilmenite raw material were detected, and all detection data were accurately entered into an Excel spreadsheet to establish a raw material component database. S2. Establishment of Acidolysis Reaction Heat Guidance Model: A parameter database was established by compiling basic data on the standard molar enthalpy of inorganic oxides, and a calculation model for the acidolysis reaction heat of ilmenite was established based on the raw material ratio. S3. Establishment of Intelligent Ore Blending Economic Model: With the minimum cost per ton of titanium as the optimization objective, the following linear programming model was constructed: min C = Σ(a i ×C i In the formula: C: the comprehensive cost of titanium product per unit after ore blending, in yuan / ton of titanium; a i The proportion of the i-th type of ilmenite (0≤x) i ≤1); C i : Unit titanium production cost when using only the i-th type of ilmenite raw material; S4, Multi-dimensional constraint setting: Delineate all constraint boundaries by combining production process indicators, impurity control, reaction conditions, and material usage ratios: Main grade constraint: Σ(a i ×TiO 2， i)≥44%~46%, impurity upper limit constraint: Σ(a i ×V2O 5， i)≤0.2%~0.3%, Σ(a i ×P2O 5， i)≤0.1%~0.15%, Σ(a i ×MnO ， i)≤0.5%~1.5%, Σ(a i ×Cr2O 3， i)≤0.1%~0.15%, Σ(a i ×Nb2O 5， i) ≤0.01%~0.05%, thermal constraint for acidolysis reaction: 800 MJ / t~880 MJ / t≤Q 混 ≤900MJ / t~1000 MJ / t; Specific mineral ratio limits: a7≤0.05-0.1, a8≤0.05-0.1, a9≤0.3-0.5; S5, Call the built-in nonlinear programming solver in Excel, use the established economic model as the objective function, set various indicators as constraints, input the relevant solver parameters and start the calculation, and finally output the optimal mineral blending scheme that meets all process constraints and the corresponding lowest production cost per ton of titanium.

2. The intelligent optimization method for the ore blending section in sulfuric acid process titanium dioxide production as described in claim 1, characterized in that: In S1, X-ray fluorescence spectroscopy was used to detect the mass content of chemical components TiO2, Fe2O3, FeO, SiO2, Al2O3, CaO, MgO, S, P2O5, Nb2O5, MnO, V2O5, and Cr2O3 in each type of ilmenite raw material.

3. The intelligent optimization method for the ore blending section in sulfuric acid process titanium dioxide production as described in claim 1, characterized in that: The calculation model for the mass fraction of each chemical component in the mixed ore established in S1 is as follows: W TiO2 =Σ(a i ×w TiO2 ,i),W Fe2O3 =Σ(a i ×w Fe2O3， i), W FeO =Σ(a i ×w FeO， i), W Al2O3 =Σ(a i ×w Al2O3， i), W CaO =Σ(a i ×w CaO， i), W S =Σ(a i ×w S， i), W P2O5 =Σ(a i ×w P2O5， i), W Nb2O5 =Σ(a i ×w Nb2O5， i), W MnO =Σa i ×w MnO， i), W V2O5 =Σ(a i ×w V2O5， i), W Cr2O3 =Σ(a i ×w Cr2O3， i); where: W represents the mass fraction of the corresponding component in the mixture; a i This represents the mass ratio of the i-th ilmenite raw material in the mixed ore; w*,i represents the original mass fraction of the corresponding component in the i-th type of ilmenite raw material; i represents the raw material ore type number.

4. The intelligent optimization method for the ore blending section in sulfuric acid process titanium dioxide production as described in claim 1, characterized in that: The guiding model for the heat of reaction of ilmenite in S2 is as follows: Heat of reaction of single ore Q i =∑(w k,i ÷M k ×ΔrH m,k ), the combined reaction heat of the mixed ore Q 混 =Σ(x i ×Q i In the formula: Q 混 The heat of reaction per unit mass of acid hydrolysis of mixed ilmenite, MJ / t; Q i w represents the heat of acidolysis per unit mass of the i-th type of single ore; k,i Content of type k oxide component in type i ore; ΔrH m,k M represents the standard molar enthalpy of reaction of the corresponding oxide; k This represents the relative molecular mass of the corresponding oxide.

Images