Intelligent optimization control method and system for alumina roasting furnace

Abstract

The invention discloses an intelligent optimization control method and system for an alumina roasting furnace, which optimizes and controls the roasting temperature based on the material balance and energy balance of the material feeding amount, the fuel amount and the output coordination; adopts a material balance algorithm to comprehensively consider the ash return amount and the feeding amount, and ensures the conservation of materials in the burner by adjusting the feeding amount; calculates the required fuel amount in real time by the energy balance algorithm to ensure the conservation of the fuel amount and the material, and ensures the stability of the roasting furnace temperature; based on the relationship between output and fuel, coordinates the total output to meet the standard by adjusting the balance between the fuel energy and the feeding amount; a roasting furnace thermal burn-off rate soft measurement model based on data analysis and a reasonable roasting temperature control point optimization method based on data analysis are established, a mathematical model of the thermal burn-off rate is established, and the roasting temperature control point is automatically optimized; through the self-optimization advance and retreat method and rolling optimization, the optimal oxygen control point with the lowest energy consumption is found to ensure the full combustion of the fuel while reducing the smoke exhaust loss.

Description

Technical Field

[0001] The present invention relates to the technical field of production automation, and in particular to an intelligent optimization control method and system for an alumina roasting furnace. Background Art

[0002] The alumina roasting system is mainly used to process qualified aluminum hydroxide sent from the flat plate. After being weighed by the belt scale, it is sent to the Venturi dryer and cyclone preheater for preheating and separation through a screw conveyor. The aluminum hydroxide with most of the crystal water removed enters the roasting furnace. The fuel (natural gas) and combustion-supporting air also enter the burner in the roasting furnace to be fully mixed and burned to release heat, and are fully mixed and contacted with the aluminum hydroxide. The temperature is about 950-1100℃. The material removes the remaining crystal water at high temperature and completes the crystal transformation. The roasted alumina enters the heat separation cyclone driven by the hot air flow and enters the four-stage cooling system for cooling. The temperature drops to about 180℃. The material enters the secondary fluidized bed cooler through the dividing valve to continue cooling. It mainly performs reverse heat exchange between the water in the internal heat exchange tube bundle and the alumina outside the tube. The alumina that comes out is cooled to below 80℃ and is transported to the alumina warehouse through the wind-driven flow trough.

[0003] At present, most of the roasting furnace control systems in the domestic alumina industry use foreign brands, such as the American Rockwell ABPLC system, the American Honeywell (China) Co., Ltd.-Honeywell DCS system, the American Emerson Electric Co., Ltd. Deltav control system, the German Siemens DCS control system, etc. Some alumina companies use domestic brands, such as Hollysys's MACSV DCS control system. These control systems can meet the basic needs of roasting furnace production. It is difficult to achieve automatic roasting in most alumina roasting furnace controls. The main circuits of the device are manually operated, the workers operate frequently, and the working conditions fluctuate greatly, resulting in the production intensity, product quality, energy consumption and material consumption of the device not reaching the optimal operating state, and the problem of benefit loss is serious. There are problems such as energy waste, environmental emissions, and insufficient intelligence level. The following main problems exist on site:

[0004] 1) Manual operation brings about unstable and unsafe production: Manual operation is greatly interfered by various factors. Even the best operators find it difficult to concentrate on observation, calculation, judgment and precise operation for a long time. Manual operation makes it difficult to keep the production equipment running and in the best condition.

[0005] 2) Problems with precise temperature control: Faced with changes in load unloading, changes in return ash, changes in fuel calorific value, frequent fluctuations in fuel pressure, and the mutual influence of multiple variables, operators are unable to comprehensively consider various factors to achieve precise digital adjustments; since the roasting temperature cannot be automatically and stably controlled, it is impossible to form a closed loop between the ignition reduction test value and the roasting temperature control, resulting in large fluctuations in the product thermal ignition reduction rate.

[0006] 3) Economic issues: Due to the different experience and technical levels of operators, each team has different operating levels for the same equipment, and even the operating methods are not completely consistent. The repeatability of the operating steps cannot be guaranteed, and the operating results are naturally different. Safety and stability are difficult to maintain for a long period of time. There is a lack of experience and means to find the best economic operating point, and production economy is difficult to guarantee.

[0007] The existing alumina industry roasting furnace control system mainly includes the following contents:

[0008] Output control: The output of the roasting furnace depends on the amount of aluminum hydroxide fed and the moisture content. The feeding rate is controlled by an electronic quantitative feeder (Schenck scale) equipped with a speed control device. The output is set by the operator on the DCS operation station. The material level and weighing sensor of the feed bin are used to adjust the feeding speed of the filtering process outside the boundary area. The operator can judge the amount of material in the bin based on this.

[0009] Product quality: Product specifications are controlled by the roasting furnace outlet temperature. The operator selects the temperature setting value based on the burning loss. The difference between the measured temperature and the set temperature is used to form a cascade control loop to control the fuel flow of the main furnace combustion station, in which the main and auxiliary controls that constitute Sichuan are both controlled by conventional PID.

[0010] Combustion efficiency: Combustion efficiency depends on the excess air at the furnace outlet or the air flow required for combustion. The air flow is achieved by changing the exhaust fan damper opening according to the oxygen content of the flue gas at the furnace outlet.

[0011] Disadvantage 1 of the prior art: The periodic change of the return ash causes the change of the roasting material (total amount and components) in the furnace. Due to the hysteresis of the combustion reaction and the hysteresis of the artificial adjustment of the fuel, the roasting temperature fluctuates widely, which has a certain impact on the product quality, and it is even more impossible to establish a more accurate relationship between the roasting temperature and the ignition loss rate;

[0012] Disadvantage 2 of the existing technology: The setting of the roasting temperature is basically based on manual experience and test data. The instability of the furnace temperature control increases the possibility of accurately obtaining the roasting temperature. It is impossible to form a closed loop between the test value of the ignition loss rate and the control of the roasting temperature, resulting in large fluctuations in the thermal ignition loss rate of the product. How to establish the relationship between the roasting temperature, the moisture content of the raw material and the ignition loss rate is crucial to improving product quality and realizing unmanned operation;

[0013] Disadvantage 3 of the existing technology: The combustion atmosphere of the roasting furnace is completely manually operated according to the oxygen content table. First, manual operation cannot meet the changing working conditions on site. Not only is the labor intensity high, but there is also energy waste caused by untimely adjustment. Second, the oxygen content is a representation of combustion, and it should be different under different working conditions. All of them are controlled step by step according to fixed values, which cannot meet the changing working conditions. Summary of the invention

[0014] To this end, the present invention provides an intelligent optimization control method and system for an alumina roasting furnace to solve the above technical problems existing in the control systems of the existing alumina roasting furnaces in the industry.

[0015] In order to achieve the above object, the present invention provides the following technical solutions:

[0016] According to a first aspect of an embodiment of the present invention, an intelligent optimization control method for an alumina roasting furnace is provided, the method comprising:

[0017] Establish a material balance model for ash return and unloading, and obtain the unloading amount control value according to the ash return pressure; establish an oxygen protection model, implement closed-loop control of oxygen according to the deviation between the oxygen setting value and the measured value, and obtain the oxygen protection analog output value; implement closed-loop control of the roasting temperature according to the deviation between the set temperature and the actual temperature of the roasting furnace, and obtain the roasting temperature analog output value; comprehensively consider the unloading amount control value, oxygen protection analog output value, roasting temperature analog output value, and the set upper and lower limits of the unloading amount, obtain the unloading amount output value of the unloading machine, and realize unloading amount control;

[0018] The real-time output setting value of the combustion furnace is calculated according to the established real-time output setting calculation model, and the closed-loop control of the output is realized according to the deviation between the real-time output setting value and the real-time output measurement value, and the output simulation output value is obtained; the unit consumption of the product is calculated, and the calculated unit consumption is input into the established output and fuel energy balance model, and the fuel demand is output; the fuel quantity output value is obtained according to the obtained output simulation output value and the fuel demand, so as to realize the fuel quantity control.

[0019] Furthermore, the method further comprises:

[0020] The oxygen content is adjusted by adjusting the combustion-supporting air volume entering the roasting furnace. The closed-loop control of the oxygen content is realized according to the deviation between the set value and the measured value of the oxygen content, and the analog output value AV of the oxygen content adjustment is obtained. 氧量调节 ;

[0021] According to the change in oxygen demand caused by the change in the amount of material discharged required by production, the analog output value of oxygen for material discharge feedforward AV is calculated. 下料前馈 , specifically:

[0022] AV 下料前馈 =F 下料量增量 *K 下料引风调节 ,

[0023] Among them, F 下料量增量 =F 下料设定K -F 下料设定K-1 , F 下料设定K-1K is the material loading setting of the previous scanning cycle. 下料引风调节 The calculation of the values ​​is based on field data analysis and data modeling;

[0024] According to the oxygen content, the analog output value AV is adjusted 氧量调节 And the oxygen content analog output value AV of the material feedforward 下料前馈 The control value of the induced draft fan valve is obtained to realize the combustion air volume control, specifically:

[0025] The general instruction of the induced draft fan valve is: AV 引风总 =AV 氧量调节 +AV 下料前馈 .

[0026] Furthermore, the method further comprises:

[0027] Obtain the historical statistical data of the test loss on ignition rate and roasting temperature, and establish a relationship model between roasting temperature and loss on ignition rate based on data analysis and machine learning methods;

[0028] According to the relationship model between the roasting temperature and the ignition reduction rate, a roasting temperature value matching the best ignition reduction rate is obtained as the set temperature of the roasting furnace.

[0029] Furthermore, a material balance model of ash return and material discharge is established, and the material discharge amount control value is obtained according to the ash return pressure, which specifically includes:

[0030] By obtaining the return ash pressure exceeding the preset value, the current return ash pressure increment and the feed amount increment when the roasting temperature is similar are found. Assuming T 焙烧1 ΔP 返灰1 , ΔF 下料1 , T 焙烧2 ΔP 返灰2 , ΔF 下料2 , ... T 焙烧n ΔP 返灰n , ΔF 下料n ;

[0031] Through the least squares modeling, the model relationship between the increment of material feeding amount and the increment of ash return pressure is obtained as follows:

[0032] ΔF 下料 =f(ΔP 返灰 )

[0033] Among them, ΔP 返灰 is the change in ash return pressure between two adjacent sampling periods, that is, P 返灰k -P 返灰k-1 ;

[0034] The output of the ash return and material balance model is the sum of the increments at all times during the commissioning phase:

[0035] F 下料 =F 下料 +ΔF 下料 ;

[0036] The feeding amount control value, the oxygen protection analog output value, the roasting temperature analog output value and the set upper and lower limits of the feeding amount are combined to obtain the feeding amount output value of the feeding machine to achieve feeding amount control, specifically including:

[0037] The total output of the blanking machine is: AV 下料总 =AV 焙烧温度 +AV 氧量保护 +F 下料 ; where F 下料 is the feeding amount control value, AV 焙烧温度 AV is the analog output value of the baking temperature. 氧量保护 Analog output value for oxygen protection;

[0038] If the output of the feeder exceeds the set upper limit of the feed amount, the output of the feeder is the upper limit value. If the output of the feeder exceeds the set lower limit of the feed amount, the output of the feeder is the lower limit value.

[0039] Furthermore, the real-time output setting value of the combustion furnace is calculated according to the established real-time output setting calculation model, specifically including:

[0040] Set the daily output to W 日 , the current cumulative output is W 已生产 , the cumulative time is T, the unit is hours, and T<24, the real-time unloading amount is F 瞬时 , unit t / h, the production output to be produced is W 待 , then the real-time production setting calculation model F sp for:

[0041] W 已生产 =W 已生产 +F 瞬时 / 7200

[0042] W 待 =W 日 -W 已生产

[0043] Fsp=W 待 / (24-T).

[0044] Furthermore, the unit consumption of the product is calculated, and the calculated unit consumption is input into the established production and fuel energy balance model to output the fuel demand, specifically including:

[0045] The product unit consumption calculation module adopts the average value DH_avg of the unit consumption in the preset time period, and the calculation method is: the ratio of the moving average value of the fuel in the preset time period to the moving average value of the output in the past two hours;

[0046] F 产量移动平均值 =f(F 时刻1 , F 时刻2 , F 时刻3 , ... F 时刻N )

[0047] F 燃料量移动平均值 =f(F 时刻1 , F 时刻2 , F 时刻3 , ... F 时刻N )

[0048] DH_avg = F 燃料量移动平均值 / F 产量移动平均值 .

[0049] Furthermore, the unit consumption of the product is calculated, and the calculated unit consumption is input into the established production and fuel energy balance model to output the fuel demand, specifically including:

[0050] The relationship between fuel demand and output is:

[0051] ΔF 燃料 =DH_avg*ΔF 产量设定

[0052] Where, ΔF 产量设定 is the change in the output setting value between two adjacent sampling periods, that is, F 产量设定k -F 产量设定k-1 ;

[0053] The output of the production and fuel energy balance model is the sum of the increments at all times during the commissioning phase:

[0054] F 燃料 =F 燃料+ ΔF 燃料 ;

[0055] The fuel quantity output value is obtained according to the obtained output value of the production simulation and the fuel demand, and the fuel quantity control is realized, which specifically includes:

[0056] The total output of fuel quantity is:

[0057] AV 燃料总 =AV 产量燃料 +F 燃料

[0058] Among them, AV 产量燃料 is the output value of the analog quantity, F 燃料 The demand for fuel.

[0059] Furthermore, the method further comprises:

[0060] The best oxygen control point is obtained as the oxygen setting value by a self-optimization algorithm; the self-optimization algorithm includes:

[0061] First, set the optimization step size SOP = ΔF A =γ*F A , where F A is the current oxygen control point, with an allowable error of ε, and the counter n = 0;

[0062] ① Constant gas flow rate setting, output the current oxygen control point as the oxygen set value to the oxygen regulation module, set the optimization controller operation flag to ON, and record an optimization objective function value J1 after a period of time;

[0063] ②Choose to increase the oxygen control point, set n=n+1, the oxygen set value increment is the set step length n*SOP, output the oxygen set value to the oxygen adjustment module; modify the current oxygen control point and turn to step ⑧;

[0064] ③If J 2 >J 1 When J 2 Assign the value of 1 , turn to step ②; if J 2 <J 1 When , determine whether it is the first optimization, that is, whether n is 1. If n = 1, it means that the search direction is wrong, and turn to step ④;

[0065] If n≠1, and |J 2 -J 1 |<ε, the current oxygen control point is in the optimal state, the optimization is over, and the optimization controller operation flag is set to OFF; if n≠1, and |J 2 -J 1 |>ε, change the step length to find SOP=-0.25*SOP and go to step ②;

[0066] ④Choose to increase the oxygen control point, set n=n+1, the oxygen set value increment is the set step length n*(-SOP), output the oxygen set value to the oxygen adjustment module; modify the current oxygen control point and turn to step ⑧;

[0067] ⑤If J 2 >J 1 When J 2 Assign the value of 1 , turn to step ④;

[0068] ⑥If J 2 <J 1 When |J 2 -J 1 |<ε, the current oxygen control point is in the optimal state, the optimization is completed, and the optimization controller operation flag is set to OFF;

[0069] ⑦If J 2 <J 1 When |J 2 -J 1 |>ε, change the step length to find SOP=-0.25*SOP and go to step ②;

[0070] ⑧ Calculation of objective function value:

[0071] The objective function value J is calculated according to the optimization objective function based on the steady-state real-time process measurement value after dynamic response. 2 That is, the calculation starts after a period of time after the optimized output acts on the device, and the time depends on the dynamic response time of the process; after the calculation, it returns to the original rotated position.

[0072] According to a second aspect of an embodiment of the present invention, an intelligent optimization control system for an alumina roasting furnace is provided, the system comprising:

[0073] The unloading amount control module is used to establish a material balance model of ash return and unloading, and obtain the unloading amount control value according to the ash return pressure; establish an oxygen protection model, implement closed-loop control of oxygen according to the deviation between the oxygen setting value and the measured value, and obtain the oxygen protection analog output value; implement closed-loop control of the roasting temperature according to the deviation between the set temperature and the actual temperature of the roasting furnace, and obtain the roasting temperature analog output value; comprehensively consider the unloading amount control value, the oxygen protection analog output value, the roasting temperature analog output value, and the set upper and lower limits of the unloading amount, obtain the unloading amount output value of the unloading machine, and realize the unloading amount control;

[0074] The fuel quantity control module is used to calculate the real-time output setting value of the combustion furnace according to the established real-time output setting calculation model, realize closed-loop control of the output according to the deviation between the real-time output setting value and the real-time output measurement value, and obtain the output simulation output value; calculate the unit consumption of the product, input the calculated unit consumption into the established output and fuel energy balance model, and output the fuel demand; obtain the fuel quantity output value according to the obtained output simulation output value and the fuel demand, and realize fuel quantity control.

[0075] Furthermore, the system further comprises:

[0076] The combustion air volume control module is used to adjust the oxygen content by adjusting the combustion air volume entering the roasting furnace, realize the closed-loop control of the oxygen content according to the deviation between the oxygen content set value and the measured value, and obtain the oxygen content adjustment analog output value AV 氧量调节 ;

[0077] According to the change in oxygen demand caused by the change in the amount of material discharged required by production, the analog output value of oxygen for material discharge feedforward AV is calculated. 下料前馈 , specifically:

[0078] AV 下料前馈 =F 下料量增量 *K 下料引风调节 ,

[0079] Among them, F 下料量增量 =F 下料设定K -F 下料设定K-1 , F 下料设定K-1 K is the material loading setting of the previous scanning cycle. 下料引风调节 The calculation of the values ​​is based on field data analysis and data modeling;

[0080] According to the oxygen content, the analog output value AV is adjusted 氧量调节 And the oxygen content analog output value AV of the material feedforward 下料前馈 The control value of the induced draft fan valve is obtained to realize the combustion air volume control, specifically:

[0081] The general instruction of the induced draft fan valve is: AV 引风总 =AV 氧量调节 +AV 下料前馈 .

[0082] The present invention has the following advantages:

[0083] The optimized control method of roasting temperature based on material balance and energy balance, material quantity, fuel quantity and output coordination is realized; firstly, the material balance algorithm is adopted to comprehensively consider the ash return amount and the material quantity, and the conservation of materials in the burner is ensured by adjusting the form of the material quantity; secondly, the required fuel quantity is calculated in real time through the energy balance algorithm to ensure the conservation of fuel quantity and material and the stability of the roasting furnace temperature; then, based on the relationship between output and fuel, the total output is coordinated to meet the standard by adjusting the balance between the energy of the fuel and the material quantity;

[0084] Through the roasting furnace thermal loss rate soft measurement model based on data analysis and the reasonable roasting temperature control point optimization method based on data analysis. After the loss rate test value is reported to the control room, the operator can manually input the value. The system establishes a mathematical model of the thermal loss rate based on historical data and manually input loss rate, as well as the collected feed amount, ash return amount, roasting temperature, fuel amount, and combustion air volume through the machine learning algorithm of data analysis, and automatically optimizes the roasting temperature control point (set value), realizing the function of floating the optimal control point on the basis of automation;

[0085] The best combustion oxygen control point is optimized by using the self-optimization algorithm and rolling optimization. This technology uses the self-optimization advance and retreat method and rolling optimization to find the best oxygen control point with the lowest energy consumption, ensuring full combustion of fuel while reducing exhaust loss.

[0086] The problem of periodic fluctuation of the main furnace temperature of the roasting furnace when the roasting furnace returns ash is solved; the problem of large fluctuation of the product ignition rate caused by large temperature fluctuation of the main furnace is solved; the problem of unreasonable oxygen content causing suboptimal roasting furnace efficiency due to ash return from the roasting furnace is solved; the problem of suboptimal roasting furnace temperature due to failure to achieve automatic closed-loop control of the ignition rate and the roasting furnace temperature is solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0088] Figure 1 A flow chart of an intelligent optimization control method for an alumina roasting furnace provided by an embodiment of the present invention;

[0089] Figure 2 A process flow chart of an alumina roasting furnace in an intelligent optimization control method for an alumina roasting furnace provided in an embodiment of the present invention;

[0090] Figure 3 A logic block diagram of material feeding control in an intelligent optimization control method for an alumina roasting furnace provided in an embodiment of the present invention;

[0091] Figure 4 A logic block diagram of fuel quantity control in an intelligent optimization control method for an alumina roasting furnace provided in an embodiment of the present invention;

[0092] Figure 5 A logic block diagram of combustion-supporting air volume control in an intelligent optimization control method for an alumina roasting furnace provided in an embodiment of the present invention;

[0093] Figure 6 A graph showing the relationship between roasting temperature and ignition loss rate in an intelligent optimization control method for an alumina roasting furnace provided in an embodiment of the present invention;

[0094] Figure 7 An optimization strategy diagram of a roasting temperature control point in an intelligent optimization control method for an alumina roasting furnace provided in an embodiment of the present invention;

[0095] Figure 8 A flow chart of a self-optimization algorithm for oxygen content in an intelligent optimization control method for an alumina roasting furnace provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0096] The following is a description of the implementation of the present invention by specific embodiments. People familiar with the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. Obviously, the described embodiments are 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 creative work are within the scope of protection of the present invention.

[0097] like Figure 1 As shown, the embodiment of the present invention proposes an intelligent optimization control method for an alumina roasting furnace, the method comprising:

[0098] S100, establish a material balance model for ash return and unloading, and obtain the unloading amount control value according to the ash return pressure; establish an oxygen protection model, implement closed-loop control of the oxygen amount according to the deviation between the oxygen setting value and the measured value, and obtain the oxygen protection analog output value; implement closed-loop control of the roasting temperature according to the deviation between the set temperature and the actual temperature of the roasting furnace, and obtain the roasting temperature analog output value; comprehensively consider the unloading amount control value, the oxygen protection analog output value, the roasting temperature analog output value, and the set upper and lower limits of the unloading amount, obtain the unloading amount output value of the unloading machine, and realize the unloading amount control;

[0099] S200. Calculate the real-time output setting value of the combustion furnace according to the established real-time output setting calculation model, implement closed-loop control of the output according to the deviation between the real-time output setting value and the real-time output measurement value, and obtain the output simulation output value; calculate the unit consumption of the product, input the calculated unit consumption into the established output and fuel energy balance model, and output the fuel demand; obtain the fuel quantity output value according to the obtained output simulation output value and the fuel demand, and realize fuel quantity control.

[0100] Furthermore, the method further comprises:

[0101] S300, adjust the oxygen content by adjusting the combustion-supporting air volume entering the roasting furnace, realize closed-loop control of the oxygen content according to the deviation between the oxygen content setting value and the measured value, and obtain the oxygen content adjustment analog output value AV 氧量调节 ;

[0102] According to the change in oxygen demand caused by the change in the amount of material discharged required by production, the analog output value of oxygen for material discharge feedforward AV is calculated. 下料前馈 , specifically:

[0103] AV 下料前馈 =F 下料量增量 *K 下料引风调节 ,

[0104] Among them, F 下料量增量 =F 下料设定K -F 下料设定K-1 , F 下料设定K-1 K is the material loading setting of the previous scanning cycle. 下料引风调节 The calculation of the values ​​is based on field data analysis and data modeling;

[0105] According to the oxygen content, the analog output value AV is adjusted 氧量调节 And the oxygen content analog output value AV of the material feedforward 下料前馈 The control value of the induced draft fan valve is obtained to realize the combustion air volume control, specifically:

[0106] The general instruction of the induced draft fan valve is: AV 引风总 =AV 氧量调节 +AV 下料前馈 .

[0107] The specific implementation contents are as follows:

[0108] Alumina roasting furnace process flow Figure 2 As shown, it includes the material direction line, air or flue gas direction line. The material direction is: L01→F01→L02→A01→A02→P01→P02→P04→P03→C01→C02→C03→C04→K01A / K01B, and the gas direction is: C04→C03→C02→C01→P04→P03→P02→A02→P01→dust collection→P16→P17→P18. There is heat exchange during the movement of air and materials. Among them, the measuring point signals in the system include material unloading weighing signal, natural gas flow signal, roasting temperature signal, flue gas oxygen content signal, flue gas CO content signal, return ash pressure signal, ignition reduction rate test, actuator natural gas regulating valve, unloading valve, induced draft fan, and temperature and pressure signals in other processes.

[0109] An intelligent optimization control method for an alumina roasting furnace proposed in an embodiment of the present invention mainly includes:

[0110] Based on the material conservation principle, a balance relationship model between the ash return amount and the material feed amount is established. When the ash return amount changes, the material feed amount is adjusted in time to achieve rapid regulation. Then, according to the deviation between the control point after the roasting temperature compensation recommended by the ignition loss rate soft measurement model and the measured value, the material feed amount is adjusted through closed-loop regulation to ensure accurate control of the roasting temperature.

[0111] The method of determining the output by the amount of fuel according to the unit output situation, by statistically analyzing the unit consumption of the unit output, based on the energy balance principle, determines the amount of fuel required to meet the output, and adjusts the fuel amount through closed-loop control according to the deviation between the output setting value and the actual output, so as to achieve the consistency of the fuel amount and output;

[0112] The optimal oxygen control point is found based on the self-optimization technology, and closed-loop precise control of the oxygen content is achieved according to the fuel quantity feedforward model and the deviation between the oxygen setting value and the measured value.

[0113] 1) Feeding quantity control

[0114] First, the roasting temperature is controlled based on the amount of material fed and the amount of ash returned based on material balance. When the fuel is constant, maintaining the balance of materials can ensure a relatively stable roasting temperature. By establishing a balanced relationship model between the amount of material fed and the amount of ash returned, when the amount of ash returned changes, the amount of material fed is corrected in time to ensure the conservation of the material in the entire system, thereby ensuring the stability of the roasting temperature; then, according to the deviation between the roasting temperature feedback value and the set value, the roasting temperature is accurately adjusted through closed-loop control.

[0115] The roasting temperature control block diagram based on material balance of feed amount and ash return amount is as follows Figure 3 As shown:

[0116] ① Ash return and material feeding balance model

[0117] By obtaining the return ash pressure is too large, find the current return ash pressure increment and material feeding increment when the roasting temperature is similar, such as T 焙烧1 ΔP 返灰1 , ΔF 下料1 , T 焙烧2 ΔP 返灰2 , ΔF 下料 2, ...T 焙烧n ΔP 返灰n , ΔF 下料n ;

[0118] Through the least squares modeling, the model relationship between the increment of material feeding amount and the increment of ash return pressure is obtained as follows:

[0119] ΔF 下料 =f(ΔP 返灰 )

[0120] Among them, ΔP 返灰 is the change in ash return pressure between two adjacent sampling periods, that is, P 返灰k -P 返灰k-1 ;

[0121] The output of the ash return and material balance model is the sum of the increments at all times during the commissioning phase:

[0122] F 下料 =F 下料 +ΔF 下料 .

[0123] ②Oxygen protection model

[0124] When the return ash is too large, the oxygen content will become low. When it is lower than 1.5%, the electrostatic precipitator will trip and stop, seriously affecting production safety. When the oxygen content is too low, the amount of material discharged is reduced accordingly to protect the oxygen content and prevent it from being too low. This module takes traditional PID control as an example. The value range of P in the PID control algorithm is 200-250, the value range of I is 200-300, and the value range of D is 30-40. The output of this module is AV 氧量保护 .

[0125] ③ Control of material feeding amount based on roasting temperature control

[0126] This control is a closed-loop control that uses the amount of material to adjust the roasting temperature. According to the deviation between the set temperature and the actual temperature of the roasting furnace, the amount of material is adjusted. The data closed-loop fine stable compensation adjustment process belongs to the slow adjustment process. This module takes PID as an example. The value range of P in the PID control algorithm is 400~650, the value range of I is 150~200, and the value range of D is 50~60. The output of this module is AV 焙烧温度 .

[0127] ④Safety limit function of feeder

[0128] This function is a safety limit function, which is designed to ensure that the feeding amount of the feeder remains within this range. If the output of the feeder exceeds the upper limit set by this module, the output is the upper limit value; if the output of the feeder exceeds the lower limit set by this module, the output is the lower limit value.

[0129] The total output of the feeder is:

[0130] AV 下料总 =AV 焙烧温度 +AV 氧量保护 +F 下料 .

[0131] 2) Fuel quantity control

[0132] A control method based on the combination of the output fuel balance calculation model and the output closed-loop control based on energy balance. In the above-mentioned feeding control, the feeding amount has been used to control the roasting temperature, so the amount of fuel determines the amount of output. According to the daily output and the output of the consumed time, the unit output of the remaining time is calculated in real time. And according to the output setting and the unit consumption of the material, the fuel amount is compensated in time and quickly to achieve the purpose of rapid adjustment. Then, according to the deviation of production, the fuel amount is fine-tuned to adjust the feeding amount, and then the production requirements are met under the premise of ensuring the roasting temperature.

[0133] Fuel quantity control control block diagram Figure 4 As shown:

[0134] ① Real-time production setting calculation model

[0135] Set the daily output to W 日 , the current cumulative output is W 已生产 , the cumulative time is T (unit: hours, and T<24), and the real-time unloading amount is F 瞬时 (Unit: t / h), sampling period is 500ms, and the output to be produced is W 待 , then the real-time production setting calculation model F sp for:

[0136] W 已生产 =W 已生产 +F 瞬时 / 7200

[0137] W 待 =W 日 -W 已生产

[0138] Fsp=W 待 / (24-T).

[0139] ②Product unit consumption calculation module

[0140] The product unit consumption calculation module uses the average value of the unit consumption in the last two hours. The calculation method is: the ratio of the moving average value of fuel in the last two hours to the moving average value of output in the last two hours.

[0141] F 产量移动平均值 =f(F 时刻1 , F 时刻2 , F 时刻3 , ... F 时刻N )

[0142] F 燃料量移动平均值 =f(F 时刻1 , F 时刻2 , F 时刻3 , ... F 时刻N )

[0143] DH_avg = F 燃料量移动平均值 / F 产量移动平均值 .

[0144] ③Production and fuel energy balance model

[0145] The relationship between fuel demand and output is:

[0146] ΔF 燃料 =DH_avg*ΔF 产量设定

[0147] Where, ΔF 产量设定 is the change in the output setting value between two adjacent sampling periods, that is, F 产量设定k -F 产量设定 k -1 ;

[0148] The output of the production and fuel energy balance model is the sum of the increments at all times during the commissioning phase:

[0149] F 燃料 =F 燃料+ ΔF 燃料 .

[0150] ④Production closed-loop control model

[0151] According to the deviation between the set value and the measured value, the fuel quantity is fine-tuned to adjust the feeding quantity. This module takes PID as an example. In the PID control algorithm, the value range of P is 200-300, the value range of I is 300-400, and the value range of D is 50-60. The output of this module is AV 产量燃料 .

[0152] The total output of fuel quantity is:

[0153] AV 燃料总 =AV 产量燃料 +F 燃料

[0154] 3) Combustion air volume control

[0155] like Figure 5 As shown, the input of the loop is the oxygen content set value, the measuring instrument is zirconia, and the controller output is the induced draft fan valve opening.

[0156] ① Oxygen content SP refers to the optimized oxygen content, which is optimized by the best oxygen content control point based on self-optimization algorithm and rolling optimization.

[0157] ② Feed forward of material feeding

[0158] When the production instructions change, the induced draft fan valve opening is changed in advance to match the oxygen demand required for the change in material discharge.

[0159] AV 下料前馈 =F 下料量增量 *K 下料引风调节 ,

[0160] Among them, F 下料量增量 =F 下料设定K -F 下料设定K-1 , F 下料设定K-1 K is the material loading setting of the previous scanning cycle. 下料引风调节 The values ​​are calculated based on field data analysis and data modeling.

[0161] ③Oxygen control output

[0162] By adjusting the ventilation volume entering the roasting furnace, the oxygen content is kept stable. This module takes PID as an example. The value range of P in the PID control algorithm is 600-700, and the value range of I is 200-230. The output of the oxygen control module is AV 氧量调节 .

[0163] The general instruction of the induced draft fan valve is: AV 引风总 =AV 氧量调节 +AV 下料前馈 .

[0164] 4) Temperature compensation model for burning rate soft measurement

[0165] ① Ignition rate soft measurement model

[0166] After the loss on ignition test value is reported to the control room, the operator can manually input the value. The system will establish a mathematical model of the thermal loss on ignition rate based on historical data, the manually input loss on ignition rate and roasting temperature, and through a machine learning algorithm for data analysis.

[0167] Taking a domestic alumina production enterprise as an example, due to the above-mentioned manual operation restrictions, the main furnace temperature and ignition loss rate (shift sample test) curve for one week is as follows: Figure 6 shown.

[0168] By statistically analyzing the relationship between roasting furnace temperature and ignition reduction, it can be found that roasting temperature is inversely proportional to ignition reduction rate.

[0169] The average burning rate of one week is as follows: ZJ_avg = (ZJ_1, ZJ_2, ZJ_3....ZJ_7) / 7;

[0170] The maximum value of the burning reduction rate in one week is: ZJ_max=MAX(ZJ_1,ZJ_2,ZJ_3....ZJ_7).

[0171] The optimal burning rate is obtained through the average and maximum burning values: ZJ_YH=ZJ_avg*k31+ZJ_max*k32, where k31 and k32 are obtained based on field data analysis and data modeling.

[0172] ② Roasting temperature recommendation

[0173] The method of controlling the burning loss of alumina is mainly to control the temperature of the main furnace of the roasting furnace. After the roasting temperature can reach a considerable control accuracy, the roasting temperature control point optimization program will take effect. This function is designed to provide the most reasonable product quality on site - thermal burning loss rate. Automatically optimize the roasting temperature control point (set value). The function of floating the optimal control point is realized on the basis of automation. Control strategies such as Figure 7 shown.

[0174] By querying the corresponding relationship between the ignition reduction rate and the roasting temperature, the roasting temperature that matches the best ignition reduction rate is found, and then the control point of the roasting temperature loop is changed to achieve safe and temperature-economical operation of the roasting furnace.

[0175] 5) Self-optimization and rolling optimization model of oxygen content control point

[0176] The theoretical basis of self-optimization is the heat balance equation of gas fuel and the forward and backward self-optimization rolling optimization algorithm, and the goal is to reduce the total loss to the minimum.

[0177] By changing the oxygen control point, we can observe whether the roasting furnace efficiency is improved to find the best oxygen content control point. Figure 8 As shown:

[0178] First, set the optimization step size SOP = ΔF A =γ*F A (γ ranges from 1% to 2%), where F A is the current oxygen control point, the allowable error ε, the value range of ε is 1 to 2, and the counter n = 0

[0179] ① Constant gas flow rate setting, output the current oxygen control point as the oxygen setting value to the oxygen regulation module, set the optimization controller operation flag to ON, and record an optimization objective function value J1 after 3 to 4 minutes;

[0180] ②Choose to increase the oxygen control point, set n=n+1, the oxygen set value increment is the set step length n*SOP, output the oxygen set value to the oxygen adjustment module; modify the current oxygen control point and turn to step ⑧;

[0181] ③If J 2 >J 1 When J 2 Assign the value of1 , turn to step ②; if J 2 <J 1 When n=1, determine whether it is the first optimization, that is, whether n is 1. If n=1, it is said that the search direction is wrong, and go to step ④. If n≠1, and |J 2 -J 1 |<ε, the current oxygen control point is in the optimal state, the optimization ends, and the optimization controller operation flag is set to OFF. If n≠1, and |J 2 -J 1 When |>ε, change the step size to set SOP=-0.25*SOP and go to step ②.

[0182] ④Choose to increase the oxygen control point, set n=n+1, the oxygen set value increment is the set step length n*(-SOP), output the oxygen set value to the oxygen adjustment module; modify the current oxygen control point and turn to step ⑧;

[0183] ⑤If J 2 >J 1 When J 2 Assign the value of 1 , turn to step ④;

[0184] ⑥If J 2 <J 1 When |J 2 -J 1 |<ε, the current oxygen control point is in the optimal state, the optimization is completed, and the optimization controller operation flag is set to OFF.

[0185] ⑦If J 2 <J 1 When |J 2 -J 1 When |>ε, change the step size to set SOP=-0.25*SOP and go to step ②.

[0186] ⑧ Calculation of objective function value:

[0187] The objective function value J is calculated according to the optimization objective function based on the steady-state real-time process measurement value after dynamic response. 2 That is, the calculation starts after a period of time after the optimized output acts on the device. The time depends on the dynamic response time of the process. For a roasting furnace, it is generally 3 to 4 minutes. Return to the original rotated position.

[0188] An intelligent optimization control method for an alumina roasting furnace proposed in an embodiment of the present invention:

[0189] By establishing a model relationship between the ash return amount and the feed amount based on material balance, the material in the roasting furnace is kept in a balanced state by adjusting the feed amount, so as to ensure a high stability of the roasting temperature;

[0190] According to the tested ignition loss rate, roasting temperature, raw material moisture, fuel quantity and oxygen content, the relationship between the ignition loss rate and the above-mentioned working condition change model is completed based on data analysis and machine learning methods; after the ignition loss rate model is established, the required roasting temperature under the ignition loss rate standard is determined by comparing with the current working conditions, and used as the roasting furnace temperature setting value to achieve automatic setting and matching of roasting temperatures under different working conditions;

[0191] Based on the material conservation principle, the material unloading quantity is automatically set according to the daily production demand, the produced output and the remaining output;

[0192] Based on the energy balance principle, the automatic matching of the feeding amount and the fuel amount is achieved through the balance relationship between the feeding amount and the fuel according to the different feeding amounts.

[0193] The above control is based on automatic compensation of the model, and closed-loop control is achieved through measurement feedback to achieve high consistency in return material, output, roasting temperature and fuel quantity;

[0194] Combustion efficiency control is based on self-optimization and rolling optimization technology to find the best oxygen content control point (floating control point) to ensure sufficient roasting while reducing smoke loss, ensuring maximum fuel utilization and achieving the goal of saving fuel;

[0195] By realizing fully intelligent optimization and automatic control of the roasting furnace, the overall stable, safe and economical operation of the roasting furnace is achieved, and the long-term operation of the roasting furnace production is guaranteed; the roasting furnace can be operated quasi-unmanned, which improves the intelligence of production and reduces the labor intensity of workers.

[0196] Corresponding to the above embodiment, this embodiment proposes an intelligent optimization control system for an alumina roasting furnace, the system comprising:

[0197] The unloading amount control module is used to establish a material balance model of ash return and unloading, and obtain the unloading amount control value according to the ash return pressure; establish an oxygen protection model, implement closed-loop control of oxygen according to the deviation between the oxygen setting value and the measured value, and obtain the oxygen protection analog output value; implement closed-loop control of the roasting temperature according to the deviation between the set temperature and the actual temperature of the roasting furnace, and obtain the roasting temperature analog output value; comprehensively consider the unloading amount control value, the oxygen protection analog output value, the roasting temperature analog output value, and the set upper and lower limits of the unloading amount, obtain the unloading amount output value of the unloading machine, and realize the unloading amount control;

[0198] The fuel quantity control module is used to calculate the real-time output setting value of the combustion furnace according to the established real-time output setting calculation model, realize closed-loop control of the output according to the deviation between the real-time output setting value and the real-time output measurement value, and obtain the output simulation output value; calculate the unit consumption of the product, input the calculated unit consumption into the established output and fuel energy balance model, and output the fuel demand; obtain the fuel quantity output value according to the obtained output simulation output value and the fuel demand, and realize fuel quantity control.

[0199] Furthermore, the system further comprises:

[0200] The combustion air volume control module is used to adjust the oxygen content by adjusting the combustion air volume entering the roasting furnace, realize the closed-loop control of the oxygen content according to the deviation between the oxygen content set value and the measured value, and obtain the oxygen content adjustment analog output value AV 氧量调节 ;

[0201] According to the change in oxygen demand caused by the change in the amount of material discharged required by production, the analog output value of oxygen for material discharge feedforward AV is calculated. 下料前馈 , specifically:

[0202] AV 下料前馈 =F 下料量增量 *K 下料引风调节 ,

[0203] Among them, F 下料量增量 =F 下料设定K -F 下料设定K-1 , F 下料设定K-1 K is the material loading setting of the previous scanning cycle. 下料引风调节 The calculation of the values ​​is based on field data analysis and data modeling;

[0204] The analog output value AV is adjusted according to the oxygen content 氧量调节 And the oxygen content analog output value AV of the material feedforward 下料前馈 The control value of the induced draft fan valve is obtained to realize the combustion air volume control, specifically:

[0205] The general instruction of the induced draft fan valve is: AV 引风总 =AV 氧量调节 +AV 下料前馈 .

[0206] The functions performed by the various components in the intelligent optimization control system for an alumina roasting furnace provided by the embodiment of the present invention have been described in detail in the above embodiments, so they will not be described in detail here.

[0207] Although the present invention has been described in detail above by general description and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.

Claims

1. An intelligent optimization control method for an alumina roasting furnace, characterized in that: The method comprises: Establish a material balance model for ash return and unloading, and obtain the unloading amount control value according to the ash return pressure; establish an oxygen protection model, implement closed-loop control of oxygen according to the deviation between the oxygen setting value and the measured value, and obtain the oxygen protection analog output value; implement closed-loop control of the roasting temperature according to the deviation between the set temperature and the actual temperature of the roasting furnace, and obtain the roasting temperature analog output value; comprehensively consider the unloading amount control value, oxygen protection analog output value, roasting temperature analog output value, and the set upper and lower limits of the unloading amount, obtain the unloading amount output value of the unloading machine, and realize unloading amount control; The real-time output setting value of the combustion furnace is calculated according to the established real-time output setting calculation model, and the closed-loop control of the output is realized according to the deviation between the real-time output setting value and the real-time output measurement value, and the output simulation output value is obtained; the unit consumption of the product is calculated, and the calculated unit consumption is input into the established output and fuel energy balance model, and the fuel demand is output; the fuel quantity output value is obtained according to the obtained output simulation output value and the fuel demand, so as to realize the fuel quantity control; Establish a material balance model for ash return and material discharge, and obtain the material discharge control value according to the ash return pressure, including: By obtaining the return ash pressure exceeding the preset value, the current return ash pressure increment and the feed amount increment when the roasting temperature is similar are found. Assuming T 焙烧1 ΔP 返灰1 , ΔF 下料1 , T 焙烧2 ΔP 返灰2 , ΔF 下料2 , ... T 焙烧n ΔP 返灰n , ΔF 下料n ; Through the least squares modeling, the model relationship between the increment of material feeding amount and the increment of ash return pressure is obtained as follows: ΔF 下料 =f(ΔP 返灰 ) Among them, ΔP 返灰 is the change in ash return pressure between two adjacent sampling periods, that is, P 返灰k -P 返灰k-1 ; The output of the ash return and material balance model is the sum of the increments at all times during the commissioning phase: F 下料 =F 下料 +ΔF 下料 ; The feeding amount control value, the oxygen protection analog output value, the roasting temperature analog output value and the set upper and lower limits of the feeding amount are combined to obtain the feeding amount output value of the feeding machine to achieve feeding amount control, specifically including: The total output of the blanking machine is: AV 下料总 =AV 焙烧温度 +AV 氧量保护 +F 下料 ; where F 下料 is the feeding amount control value, AV 焙烧温度 AV is the analog output value of the baking temperature. 氧量保护 Analog output value for oxygen protection; If the output of the feeder exceeds the set upper limit of the feeder amount, the output of the feeder is the upper limit value; if the output of the feeder exceeds the set lower limit of the feeder amount, the output of the feeder is the lower limit value; The real-time output setting value of the combustion furnace is calculated according to the established real-time output setting calculation model, which specifically includes: Set the daily output to W 日 , the current cumulative output is W 已生产 , the cumulative time is T, the unit is hours, and T<24, the real-time unloading amount is F 瞬时 , unit t / h, the production output to be produced is W 待 , then the real-time production setting calculation model F sp for: W 已生产 =W 已生产 +F 瞬时 / 7200 IN 待 =In 日 -IN 已生产 Fsp=W 待 / (24-T); Calculate the unit consumption of the product, input the calculated unit consumption into the established production and fuel energy balance model, and output the fuel demand, including: The product unit consumption calculation module adopts the average value DH_avg of the unit consumption in the preset time period, and the calculation method is: the ratio of the moving average value of the fuel in the preset time period to the moving average value of the output in the past two hours; F 产量移动平均值 =f(F 时刻1 ,F 时刻2 ,F 时刻3 ,……F 时刻N ) F 燃料量移动平均值 =f(F 时刻1 ,F 时刻2 ,F 时刻3 ,……F 时刻N ) DH_avg=F 燃料量移动平均值 / F 产量移动平均值 ; Calculate the unit consumption of the product, input the calculated unit consumption into the established production and fuel energy balance model, and output the fuel demand, including: The relationship between fuel demand and output is: ΔF 燃料 =DH_avg*ΔF 产量设定 Where, ΔF 产量设定 is the change in the output setting value between two adjacent sampling periods, that is, F 产量设定k -F 产量设定k-1 , DH_avg is the ratio of the moving average value of fuel in the past two hours to the moving average value of production in the past two hours; The output of the production and fuel energy balance model is the sum of the increments at all times during the commissioning phase: F 燃料 =F 燃料 +ΔF 燃料 ; The fuel quantity output value is obtained according to the obtained output value of the production simulation and the fuel demand, and the fuel quantity control is realized, which specifically includes: The total output of fuel quantity is: OF 燃料总 =OFF 产量燃料 +F 燃料 Among them, AV 产量燃料 is the output value of the analog quantity, F 燃料 The demand for fuel.

2. The intelligent optimization control method for an alumina roasting furnace according to claim 1, characterized in that: The method further comprises: The oxygen content is adjusted by adjusting the combustion-supporting air volume entering the roasting furnace. The closed-loop control of the oxygen content is realized according to the deviation between the set value and the measured value of the oxygen content, and the analog output value AV of the oxygen content adjustment is obtained. 氧量调节 ; According to the change in oxygen demand caused by the change in the amount of material discharged required by production, the analog output value of oxygen for material discharge feedforward AV is calculated. 下料前馈 , specifically: OF 下料前馈 =F 下料量增量 *K 下料引风调节 , Among them, F 下料量增量 =F 下料设定K -F 下料设定K-1 , F 下料设定K-1 K is the material loading setting of the previous scanning cycle. 下料引风调节 The calculation of the values ​​is based on field data analysis and data modeling; The analog output value AV is adjusted according to the oxygen content 氧量调节 And the oxygen content analog output value AV of the material feedforward 下料前馈 The control value of the induced draft fan valve is obtained to realize the combustion air volume control, specifically: The general instruction of the induced draft fan valve is: AV 引风总 =AV 氧量调节 +AV 下料前馈 .

3. The intelligent optimization control method for an alumina roasting furnace according to claim 1, characterized in that: The method further comprises: Obtain the historical statistical data of the test loss on ignition rate, material feeding amount, ash return amount, roasting temperature, fuel amount, and combustion-supporting air volume, and establish a mathematical model of thermal loss on ignition rate based on data analysis and machine learning methods; According to the mathematical model of the thermal ignition loss rate, a roasting temperature value matching the best ignition loss rate is obtained as the set temperature of the roasting furnace.

4. The intelligent optimization control method for an alumina roasting furnace according to claim 1, characterized in that: The method further comprises: The best oxygen control point is obtained as the oxygen setting value by a self-optimization algorithm; the self-optimization algorithm includes: First, set the optimization step size SOP = ΔF A =γ*F A , where F A is the current oxygen control point, the allowable error ε, the counter n = 0, and the value range of γ is 1% to 2%; ① Constant gas flow rate setting, output the current oxygen control point as the oxygen set value to the oxygen regulation module, set the optimization controller operation flag to ON, and record an optimization objective function value J1 after a period of time; ②Choose to increase the oxygen control point, set n=n+1, the oxygen set value increment is the set step length n*SOP, output the oxygen set value to the oxygen adjustment module; modify the current oxygen control point and turn to step ⑧; ③If J2>J1, it means the search direction is correct. Continue searching in this direction, assign the value of J2 to J1, and go to step ②; if J2<J1, determine whether it is the first optimization, that is, n=1. If n=1, it means the search direction is wrong, and go to step ④; If n≠1, and |J2-J1|<ε, the current oxygen control point is in the optimal state, the optimization ends, and the optimization controller operation flag is set to OFF; if n≠1, and |J2-J1|>ε, the variable step size is used to set SOP=-0.25*SOP, and go to step ②; ④Choose to increase the oxygen control point, set n=n+1, the oxygen set value increment is the set step length n*(-SOP), output the oxygen set value to the oxygen adjustment module; modify the current oxygen control point and turn to step ⑧; ⑤If J2>J1, it means the search direction is correct. Continue searching in this direction, assign the value of J2 to J1, and go to step ④; ⑥ If J2<J1, and |J2-J1|<ε, the current oxygen control point is in the optimal state, the optimization is completed, and the optimization controller operation flag is set to OFF; ⑦If J2<J1, and |J2-J1|>ε, then change the step size to set SOP=-0.25*SOP and go to step ②; ⑧ Calculation of objective function value: The objective function value J2 is calculated according to the optimization objective function using the steady-state real-time process measurement value after dynamic response, that is, the calculation starts after a period of time after the optimization output acts on the device, and the time depends on the dynamic response time of the process; after the calculation, it returns to the original rotated position.

5. An intelligent optimization control system for an alumina roasting furnace, characterized in that: The system comprises: The unloading amount control module is used to establish a material balance model of ash return and unloading, and obtain the unloading amount control value according to the ash return pressure; establish an oxygen protection model, implement closed-loop control of oxygen according to the deviation between the oxygen setting value and the measured value, and obtain the oxygen protection analog output value; implement closed-loop control of the roasting temperature according to the deviation between the set temperature and the actual temperature of the roasting furnace, and obtain the roasting temperature analog output value; comprehensively consider the unloading amount control value, the oxygen protection analog output value, the roasting temperature analog output value, and the set upper and lower limits of the unloading amount, obtain the unloading amount output value of the unloading machine, and realize the unloading amount control; The fuel quantity control module is used to calculate the real-time output setting value of the combustion furnace according to the established real-time output setting calculation model, realize the closed-loop control of the output according to the deviation between the real-time output setting value and the real-time output measurement value, and obtain the output simulation output value; calculate the unit consumption of the product, input the calculated unit consumption into the established output and fuel energy balance model, and output the fuel demand; obtain the fuel quantity output value according to the obtained output simulation output value and the fuel demand, and realize the fuel quantity control; Establish a material balance model for ash return and material discharge, and obtain the material discharge control value according to the ash return pressure, including: By obtaining the return ash pressure exceeding the preset value, the current return ash pressure increment and the feed amount increment when the roasting temperature is similar are found. Assuming T 焙烧1 ΔP 返灰1 , ΔF 下料1 , T 焙烧2 ΔP 返灰2 , ΔF 下料2 , ... T 焙烧n ΔP 返灰n , ΔF 下料n ; Through the least squares modeling, the model relationship between the increment of material feeding amount and the increment of ash return pressure is obtained as follows: ΔF 下料 =f(ΔP 返灰 ) Among them, ΔP 返灰 is the change in ash return pressure between two adjacent sampling periods, that is, P 返灰k -P 返灰k-1 ; The output of the ash return and material balance model is the sum of the increments at all times during the commissioning phase: F 下料 =F 下料 +ΔF 下料 ; The feeding amount control value, the oxygen protection analog output value, the roasting temperature analog output value and the set upper and lower limits of the feeding amount are combined to obtain the feeding amount output value of the feeding machine to achieve feeding amount control, specifically including: The total output of the blanking machine is: AV 下料总 =AV 焙烧温度 +AV 氧量保护 +F 下料 ; where F 下料 is the feeding amount control value, AV 焙烧温度 AV is the analog output value of the baking temperature. 氧量保护 Analog output value for oxygen protection; If the output of the feeder exceeds the set upper limit of the feeder amount, the output of the feeder is the upper limit value; if the output of the feeder exceeds the set lower limit of the feeder amount, the output of the feeder is the lower limit value; The real-time output setting value of the combustion furnace is calculated according to the established real-time output setting calculation model, which specifically includes: Set the daily output to W 日 , the current cumulative output is W 已生产 , the cumulative time is T, the unit is hours, and T<24, the real-time unloading amount is F 瞬时 , unit t / h, the production output to be produced is W 待 , then the real-time production setting calculation model F sp for: W 已生产 =W 已生产 +F 瞬时 / 7200 IN 待 =In 日 -IN 已生产 Fsp=W 待 / (24-T); Calculate the unit consumption of the product, input the calculated unit consumption into the established production and fuel energy balance model, and output the fuel demand, including: The product unit consumption calculation module adopts the average value DH_avg of the unit consumption in the preset time period, and the calculation method is: the ratio of the moving average value of the fuel in the preset time period to the moving average value of the output in the past two hours; F 产量移动平均值 =f(F 时刻1 ,F 时刻2 ,F 时刻3 ,……F 时刻N ) F 燃料量移动平均值 =f(F 时刻1 ,F 时刻2 ,F 时刻3 ,……F 时刻N ) DH_avg=F 燃料量移动平均值 / F 产量移动平均值 ; Calculate the unit consumption of the product, input the calculated unit consumption into the established production and fuel energy balance model, and output the fuel demand, including: The relationship between fuel demand and output is: ΔF 燃料 =DH_avg*ΔF 产量设定 Where, ΔF 产量设定 is the change in the output setting value between two adjacent sampling periods, that is, F 产量设定k -F 产量设定k-1 , DH_avg is the ratio of the moving average value of fuel in the past two hours to the moving average value of production in the past two hours; The output of the production and fuel energy balance model is the sum of the increments at all times during the commissioning phase: F 燃料 =F 燃料 +ΔF 燃料 ; The fuel quantity output value is obtained according to the obtained output value of the production simulation and the fuel demand, and the fuel quantity control is realized, which specifically includes: The total output of fuel quantity is: OF 燃料总 =OFF 产量燃料 +F 燃料 Among them, AV 产量燃料 is the output value of the analog quantity, F 燃料 The demand for fuel.

6. The intelligent optimization control system for an alumina roasting furnace according to claim 5, characterized in that: The system further comprises: The combustion air volume control module is used to adjust the oxygen content by adjusting the combustion air volume entering the roasting furnace, realize the closed-loop control of the oxygen content according to the deviation between the oxygen content set value and the measured value, and obtain the oxygen content adjustment analog output value AV 氧量调节 ; According to the change in oxygen demand caused by the change in the amount of material discharged required by production, the analog output value of oxygen for material discharge feedforward AV is calculated. 下料前馈 , specifically: OF 下料前馈 =F 下料量增量 *K 下料引风调节 , Among them, F 下料量增量 =F 下料设定K -F 下料设定K-1 , F 下料设定K-1 K is the material loading setting of the previous scanning cycle. 下料引风调节 The calculation of the values ​​is based on field data analysis and data modeling; The analog output value AV is adjusted according to the oxygen content 氧量调节 And the oxygen content analog output value AV of the material feedforward 下料前馈 The control value of the induced draft fan valve is obtained to realize the combustion air volume control, specifically: The general instruction of the induced draft fan valve is: AV 引风总 =AV 氧量调节 +AV 下料前馈 .

Images