Furnace temperature uniformity regulation and control method and system for high-temperature continuous boat-pushing sintering furnace

Abstract

The invention belongs to the technical field of push boat sintering furnace regulation and control, and particularly relates to a high-temperature continuous push boat sintering furnace temperature uniformity regulation and control method and system. Comprising the following steps that S1, a sintering furnace is divided into a pre-sintering section, a sintering section, a slow cooling section and a water cooling section according to process requirements, heating bodies are arranged on the pre-sintering section and the sintering section, and different heat preservation materials are selected for building according to heat preservation parameters of different temperature sections; s2, dividing a pre-sintering section and a sintering section into a plurality of heating areas, and smoothly distributing the heating power of each heating area by adopting an improved average power distribution method, so that the temperature of the pre-sintering section and the sintering section is in smooth transition; and S3, monitoring temperature change data information of the pre-sintering section and the sintering section, and carrying out PID parameter self-adjustment by utilizing an improved fuzzy control rule to realize smooth distribution of the furnace temperature in the hearth. The high-temperature sintering method has the advantages that the high-temperature sintering method can be used for high-temperature sintering of ceramic-grade uranium dioxide pellets and can stably operate for a long time under the working conditions of the temperature of 1750-1760 DEG C and the hydrogen atmosphere.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of push boat sintering furnace regulation and control, and particularly relates to a high-temperature continuous push boat sintering furnace temperature uniformity regulation method and system. BACKGROUND

[0002] The high-temperature continuous push boat sintering furnace is a furnace type capable of continuously completing the sintering function of ceramic UO2 pellets. Under the high temperature of 1700-1800 DEG C and the hydrogen atmosphere, the green pellets are changed into crystal combinations through the heat treatment process of sintering, so as to achieve the required physical and mechanical properties. It is an extremely important device in the production of nuclear fuel pellets.

[0003] Prior art one, invention patent: a continuous high-temperature sintering furnace for nuclear fuel pellets (application number: CN202311847753.1), the disclosed content is that a muffle is arranged in the furnace body, a molybdenum heating body is arranged outside the muffle in the furnace body, a gas inlet pipe for conveying dry hydrogen is arranged on the furnace body, the feeding chamber and the discharging chamber are both provided with furnace doors and isolation doors, the feeding chamber and the discharging chamber are also provided with a vacuum pumping system and a nitrogen supply mechanism, a sealing layer is arranged on the outer side of the furnace body, and the hearth of the furnace body includes a furnace bottom, a furnace top and a furnace wall, the furnace bottom, the furnace wall and the furnace top of the hearth adopt dry masonry and mortise and tenon structure.

[0004] Prior art two, invention patent: step beam type sintering furnace (application number: CN201611241558.4), the disclosed content is that the sintering furnace adopts a design scheme combining arching and hoisting, the high-temperature zone hot face brick is arranged in an arching manner, and the heat insulation material and the heat preservation material adopt a hoisting structure, thereby reducing the risk of furnace collapse and boat sticking; the top cover structure is optimized, the number of alumina "I" shaped bricks is reduced, the heat preservation effect of the high-temperature zone hearth is improved, and the surface temperature of the top cover of the high-temperature zone is reduced.

[0005] In summary, the design and manufacture of the existing push boat sintering furnace equipment still have some key technologies to be solved, mainly in that the existing technology has the problems of low temperature uniformity regulation precision and poor stability effect in the high-temperature sintering of the pellets. SUMMARY

[0006] The purpose of the present application is to provide a high-temperature continuous push boat sintering furnace temperature uniformity regulation method and system, which can solve the problems of low temperature uniformity regulation precision and poor stability effect in the high-temperature sintering of the pellets.

[0007] The technical scheme of the present application is as follows: a high-temperature continuous push boat sintering furnace temperature uniformity regulation method, comprising the following steps:

[0008] S1: According to the process requirements, the sintering furnace is divided into pre-sintering section, sintering section, slow cooling section and water cooling section, and heating bodies are arranged in the pre-sintering section and sintering section, and different heat preservation materials are selected according to the heat preservation parameters of different temperature sections;

[0009] S2: The pre-sintering section and the sintering section are divided into multiple heating zones, and the improved average power distribution method is used to smoothly distribute the heating power of each heating zone, so that the temperature of the pre-sintering section and the sintering section smoothly transitions;

[0010] S3: Monitor the temperature change data information of the pre-sintering section and the sintering section, and use the improved fuzzy control rule to adjust the PID parameters automatically, so as to realize the smooth distribution of the furnace temperature in the furnace.

[0011] In the step S1, the pre-sintering section and the sintering section are connected by a transition section, the slow cooling section and the water cooling section are not provided with heating bodies, and the thickness of the furnace lining of the water cooling section is less than that of the slow cooling section.

[0012] In the step S1, the furnace of the sintering section is a corundum muffle made of corundum brick materials, and the top of the furnace is wedge-shaped and arched; the back-up brick of the corundum muffle is a double-layered alumina hollow ball brick, the outer layer of which is a furnace wall made of mullite insulating bricks, and the outer layer of the furnace wall is an insulating layer composed of silicon-calcium plates.

[0013] In the step S2, the pre-sintering section is provided with three heating zones, and the heating temperature is 1100±10℃; the pre-sintering section is symmetrically heated from left to right, and the rated heating power of each heating zone is 18kW-20kW; the sintering section is provided with four heating zones, and the rated heating temperature is 1755±5℃; the sintering section is symmetrically heated from left to right, and the rated heating power of each heating zone is 45kW-50kW.

[0014] In the step S2, the improved average power distribution method is used to smoothly distribute the heating power of each heating zone, which includes: constructing a heating power network framework based on generative adversarial imitation learning, training the discrimination module and the generation module with expert data, and confusing the judgment of discrimination, and using the discrimination module to distinguish the data distribution generated by the generation module; initializing the power as P0, and the sampling trajectory as The power data sampling trajectory generated by the generation module is The sampling trajectory of the real power in a certain heating zone is;

[0015] The expression is:

[0016]

[0017] In the formula, is the function value of the power distribution mode of gradient descent update, f is the number of heating zones, and z(i) H(k) is the real data of a heating zone node, H(k (i) ) is the weight of generated data, U(H(k) is the weight determination value of real power in the sampling track of a heating zone, U(H(k (i) )) is the weight determination value of generated data, U() is the weight determination of the generated power data sampling track generated by the generation module, U(z (i) ) is the weight determination of the real data by the discrimination module;

[0018] The loss function of the generative adversarial network is constructed, and the reward function is constructed with the discrimination module, and the generation module is trained, and the expression is:

[0019]

[0020] In the formula, is the loss function value of the gradient descent update of the generative adversarial network;

[0021] When the discrimination module cannot distinguish the data generated by the generation module from the real data, the generation module and the discrimination module reach Nash equilibrium, and the generation module successfully matches the expert power distribution method;

[0022] The parameter optimization of the generation module network is performed based on the improved average power distribution method optimization method, and the power distribution uniform sample based on the improved average power distribution method optimization method is as follows:

[0023]

[0024] In the formula, F CLIP (ω) is the optimized power distribution uniform value, is the average power distribution method optimization method, min( ) is the minimum function value, e t (ω) is the heating power distribution weight of a heating zone, is the estimated value at a moment, clip( ) is the truncation threshold modification, and the heating power target function of the heating zone is replaced, and ρ is the truncation threshold.

[0025] In the step S2, the parameter optimization of the generation module network is performed based on the improved average power distribution method optimization method, including the following steps:

[0026] (1) Generalized average distribution advantage estimation

[0027] The power obtained by executing the heating action J under the stable current T is compared with the current power distribution mode P(T|J) to evaluate the appropriateness and deviation of the heating action J, and the power distribution mode is updated in a direction in which the average distribution advantage estimation is greater than zero, and the weight M(J|T) of the heating action J is updated; the generalized average distribution advantage estimation is as follows:

[0028]

[0029] In the formula, is the generalized average distribution advantage estimation value, and ∞ is infinity, (μ,η) n is a heating coefficient, and the value of the discount rate constraint is cycled n times, is a heating power state value function at the current time, W(T t+n+1 ) is a current value function of the heating state at the next time, e t+n is a power function at the current time, W(T t+n ) is a stable current value function at the current time, and η is a discount rate of the value function at the previous time;

[0030] (2) The stable current-heating action weight ratio e t (ω) is expressed as:

[0031]

[0032] In the formula, P ω (J t |T t ) is the logarithmic weight of a batch of heating action sets executed by the new power distribution mode, and P old (J t |T t ) is the logarithmic weight of the average heating action executed by the old power distribution mode.

[0033] In the step S3, the PID parameter is self-adjusted by using the improved fuzzy control rule, including: the fuzzy inference result is de-fuzzified, the inference result is converted into an accurate value, the maximum entropy weight value method is adopted, the maximum entropy weight value de-fuzzification method selects the element with the maximum entropy weight value in the fuzzy quantity of the inference result as the output value, and the expression is as follows:

[0034]

[0035] In the formula, Γ0 is the output value of the element with the maximum entropy weight value in the fuzzy quantity of the inference result, max( ) is the maximum function, is the calculation method of the maximum entropy weight value de-fuzzification, Γ is the output quantity of the maximum entropy weight value de-fuzzification, and S is the output universe;

[0036] If the maximum entropy weight value corresponds to more than one output value in the output domain S, the average value of the output with the maximum entropy weight value is taken, and the expression is:

[0037]

[0038] In the formula, I is the total number of outputs with the maximum same entropy weight value, Γ i is the value of the output at a certain entropy weight.

[0039] A high-temperature continuous push boat sintering furnace temperature uniformity control system, including temperature section division and insulation material selection module, for dividing the sintering furnace into pre-burning section, sintering section, slow cooling section, water cooling section according to process requirements, and arranging heating bodies in the pre-burning section and the sintering section, and selecting different insulation materials according to the insulation parameters of different temperature sections.

[0040] An average power distribution module is used for dividing the pre-burning section and the sintering section into multiple heating zones, and using an improved average power distribution method to smoothly distribute the heating power of each heating zone, so that the temperature of the pre-burning section and the sintering section smoothly transitions.

[0041] A temperature adjustment module is used for monitoring the temperature change data information of the pre-burning section and the sintering section, and using an improved fuzzy control rule to adjust the PID parameters automatically, so as to realize the smooth distribution of the furnace temperature in the hearth.

[0042] An electrical control system is used for taking PLC as the main control unit, collecting the process parameters of the gas supply system of each temperature section by using multiple different sensors, transmitting the signals generated by the collected process parameters to PLC and industrial computer through a circuit for monitoring, and comprehensively monitoring the operation and stable current of the high-temperature continuous push boat sintering furnace to complete the automatic setting and monitoring of the process parameters.

[0043] The beneficial effects of the present application are that the present application combines the characteristics of the resistance heating hydrogen atmosphere sintering equipment, comprehensively considers the requirements of temperature uniformity, atmosphere uniformity and heating process matching, and provides a set of convenient and practical, safe and reliable, and good performance core block high-temperature continuous push boat sintering furnace, and realizes the full-automatic operation of the whole sintering process in combination with the peripheral automation line. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1 is a high-temperature continuous push boat sintering furnace temperature uniformity regulation method flow chart provided by the embodiment of the application;

[0045] Figure 2 is a UO2 pellet sintering process curve provided by the embodiment of the application;

[0046] Figure 3 is an air inlet and outlet position diagram provided by the embodiment of the application;

[0047] Figure 4 is a high-temperature continuous push boat sintering furnace temperature uniformity regulation system schematic diagram provided by the embodiment of the application.

[0048] In the figure: 1, temperature section division and insulation material selection module; 2, average power distribution module; 3, temperature adjustment module; 4, electrical control system. DETAILED DESCRIPTION

[0049] The application will be further described in detail below in combination with the drawings and specific embodiments.

[0050] A high-temperature continuous push boat sintering furnace temperature uniformity regulation method, comprising the following steps:

[0051] S1: According to the process requirements, the sintering furnace is divided into a pre-sintering section, a sintering section, a slow cooling section and a water cooling section, and heating bodies are arranged in the pre-sintering section and the sintering section, and different insulation materials are selected for masonry according to the insulation parameters of different temperature sections;

[0052] In step S1, the pre-sintering section and the sintering section are communicated through a transition section, no heating body is arranged in the slow cooling section and the water cooling section, and the thickness of the furnace lining of the water cooling section is less than that of the slow cooling section, which meets the material discharge cooling requirement under the condition of meeting the insulation performance.

[0053] Further, the hearth of the sintering section is a corundum muffle masonry, the top of the hearth is wedge-shaped and arched, the back-up brick of the corundum muffle is a double-layered alumina hollow ball brick, the outer layer of the double-layered alumina hollow ball brick is a masonry furnace wall of mullite insulating brick, and the outer layer of the furnace wall is an adiabatic layer composed of a silicon-calcium plate.

[0054] S2: The pre-sintering section and the sintering section are divided into multiple heating zones, and an improved average power distribution method is used to smoothly distribute the heating power of each heating zone, so that the temperature of the pre-sintering section and the sintering section smoothly transitions;

[0055] In step S2, the pre-sintering section is provided with three heating zones, and the rated heating temperature is 1100±10℃, the pre-sintering section is heated in a left-right symmetrical manner, and the rated heating power arranged in each heating zone is 18kW-20kW;

[0056] The sintering section is provided with four heating zones, and the rated heating temperature is 1755±5℃. The heating mode of the sintering section is left-right symmetrical heating, and the rated heating power arranged in each heating zone is 45kW-50kW.

[0057] In step S2, the improved average power distribution method is used to smooth the heating power of each heating zone, including: constructing a heating power network framework based on generative adversarial imitation learning, training the discriminant module and the generation module with expert data, confusing the judgment of the discriminant, and using the discriminant module to distinguish the data distribution generated by the generation module; initializing the power as P0, and the sampling trajectory as The power data sampling trajectory generated by the generation module is The sampling trajectory of the real power in a certain heating zone is

[0058] The discriminant module uses the power distribution mode function for gradient descent update, and then distinguishes the weight distribution of the real data and the generated data, and the expression is:

[0059]

[0060] In the formula, is the power distribution mode function value of the gradient descent update, f is the number of heating zones, z (i) is the real data of a certain heating zone node, H(k (i) ) is the weight of the generated data, is the weight determination value of the real power in a certain heating zone in the sampling trajectory, U(H(k (i) )) is the weight determination value of the generated data, is the weight determination value in the power data sampling trajectory generated by the generation module, u() is the weight determination, and u(z (i) ) is the weight determination of the real data by the discriminant module.

[0061] The loss function of the generative adversarial network is constructed, and the reward function is constructed with the discriminant module to train the generation module, and the expression is:

[0062]

[0063] In the formula, is the loss function value of the generative adversarial network gradient descent update;

[0064] When the discriminant module cannot distinguish the data generated by the generation module from the real data, the generation module and the discriminant module reach Nash equilibrium, and the generation module successfully matches the expert power distribution mode;

[0065] The core task of the discriminator is to distinguish between real data and generated data, and its function is essentially binary classification:

[0066] (1) Input judgment: receive real data or generated data (created by the generation module), output the probability value (0-1) that the data belongs to real data.

[0067] (2) Dynamic optimization: continuously adjust its own parameters to improve the recognition accuracy of generated data, forcing the generation module to improve output quality;

[0068] The generation module includes:

[0069] (1) Basic data generation; pre-defined data template generates structured pseudo-data, supporting multi-language and regional configuration;

[0070] (2) Simulation engine; generate random numbers based on probability distribution (such as normal distribution, uniform distribution) to simulate complex system behavior.

[0071] After the generation module generates data, the generation module network is optimized based on the improved average power allocation method, and the power allocation uniform sample based on the improved average power allocation method is as follows:

[0072]

[0073] In the formula, F CLIP (ω) is the optimized power allocation uniform value, is the average power allocation method, min( ) is the minimum function value, e t (ω) is the heating power allocation weight of a heating zone, is the estimated value at a certain moment, clip( ) is the truncation threshold modification, and ρ is the truncation threshold.

[0074] Further, the generation module network is optimized based on the improved average power allocation method, including the following steps:

[0075] (1) Generalized average allocation advantage estimation

[0076] Compare the power obtained by executing the heating action J under the stable current T with the current power allocation method P(T|J), to evaluate the suitability and deviation of the heating action J, so that the power allocation method is updated to the direction where the generalized average allocation advantage estimation is greater than zero. The weight M(J|T) of the heating action J; the generalized average allocation advantage estimation is as follows:

[0077]

[0078] wherein, is a generalized average distribution advantage estimation value, ∞ is infinity, (μ,η) n is a heating coefficient, a discount rate constraint value of n times of circulation, is a heating power state value function of the current time, W(T t+n+1 ) is a current value function of the next time heating state, e t+n is a power function of the current time, W(T t+m ) is a stable current value function of the current time, η is a discount rate of the previous time value function;

[0079] (2) The stable current-heating action weight ratio e t (ω) is expressed as:

[0080]

[0081] wherein, P ω (J t |T t ) is a logarithmic weight of a certain batch of heating action set executed by the new power distribution mode, P old (J t |T t ) is an average heating action logarithmic weight executed by the old power distribution mode.

[0082] S3: Monitor the temperature change data information of the pre-burning section and the sintering section, and use the improved fuzzy control rule to adjust the PID parameters automatically to realize the smooth distribution of the furnace temperature in the furnace.

[0083] In step S3, the improved fuzzy control rule is used to adjust the PID parameters automatically, including:

[0084] The fuzzy inference result is de-fuzzified, and the inference result is converted into an accurate value. The control system uses the maximum entropy weight value method, which does not consider the shape of the output entropy weight function, but only considers the output value at the maximum entropy weight value. The maximum entropy weight de-fuzzification method selects the element with the maximum entropy weight value in the fuzzy quantity of the inference result as the output value, and the expression is:

[0085]

[0086] wherein, Γ0 is the output value of the element with the maximum entropy weight value in the fuzzy quantity of the inference result, max( ) is the maximum function, is the calculation method of the maximum entropy weight de-fuzzification, Γ is the output quantity of the maximum entropy weight de-fuzzification, S is the output universe;

[0087] If there are more than one output values corresponding to the maximum entropy weight value in the output universe S, the average value of the output with the maximum entropy weight value is taken, and the expression is:

[0088]

[0089] In the formula, I is the total number of the maximum same entropy weight output, Γ i is the value of a certain entropy weight output.

[0090] In step S1, in the water cooling section, the electrical control system with PLC as the main control unit also monitors and controls the tail gas discharge system, the electrical control system with PLC as the main control unit collects the process parameters of the gas supply system in each temperature section by using multiple different sensors, the process parameters include gas pipeline pressure, water pipeline pressure, furnace pressure, furnace temperature, dew point, water temperature, and the signals generated by the collected process parameters are transmitted to the PLC and the industrial computer through a circuit for monitoring, and the integrated industrial configuration system is uniformly controlled, the operation of the high-temperature continuous push boat sintering furnace and the stable current are monitored in all directions, and the automatic setting and monitoring of the process parameters are completed.

[0091] The gas supply system includes a hydrogen gas supply system, a nitrogen gas supply system, a natural gas supply system and a compressed air supply system.

[0092] Another object of the present application is to provide a high-temperature continuous push boat sintering furnace temperature uniformity control system, which implements the high-temperature continuous push boat sintering furnace temperature uniformity control method, and the system comprises:

[0093] A temperature section division and insulation material selection module is used to divide the sintering furnace into a pre-burning section, a sintering section, a slow cooling section and a water cooling section according to process requirements, and heating bodies are arranged in the pre-burning section and the sintering section, and different insulation materials are selected for masonry according to the insulation parameters of different temperature sections.

[0094] An average power distribution module is used to divide the pre-burning section and the sintering section into multiple heating zones, and an improved average power distribution method is used to smoothly distribute the heating power of each heating zone, so that the temperature of the pre-burning section and the sintering section smoothly transitions.

[0095] A temperature adjustment module is used to monitor the temperature change data information of the pre-burning section and the sintering section, and an improved fuzzy control rule is used for PID parameter self-adjustment to realize the smooth distribution of the furnace temperature in the hearth.

[0096] The system further comprises:

[0097] An electrical control system is used to take PLC as the main control unit, collect the process parameters of the gas supply system in each temperature section by using multiple different sensors, and transmit the signals generated by the collected process parameters to the PLC and the industrial computer through a circuit for monitoring, and the integrated industrial configuration system is uniformly controlled, the operation of the high-temperature continuous push boat sintering furnace and the stable current are monitored in all directions, and the automatic setting and monitoring of the process parameters are completed.

[0098] The electrical control system, with a PLC as the main control unit, also monitors and regulates the exhaust gas emission system.

[0099] Example 1:

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications to the technical solutions described in the foregoing embodiments, or equivalent substitutions for some or all of the technical features, do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

[0101] Example 1, such as Figure 1 As shown, the method for controlling the temperature uniformity of a high-temperature continuous push-boat sintering furnace includes:

[0102] S1. According to the process requirements, the sintering furnace is divided into a pre-firing section, a sintering section, a slow cooling section, and a water cooling section. Heating bodies are arranged in the pre-firing section and the sintering section. Different insulation materials are selected for construction according to the insulation parameters of different temperature sections.

[0103] S2, the pre-sintering section and sintering section are divided into multiple heating zones, and an improved average power distribution method is used to smoothly distribute the heating power of each heating zone, so that the temperature transition between the pre-sintering section and the sintering section is smooth.

[0104] S3 monitors temperature change data in the pre-burning and sintering sections, and uses improved fuzzy control rules to automatically adjust PID parameters, thereby achieving a smooth temperature distribution in the furnace.

[0105] For example, in step S1, the pre-burning section and the sintering section are connected by a transition section, and the slow cooling section and the water cooling section are not equipped with heating elements. The furnace lining thickness of the water cooling section is less than that of the furnace lining thickness of the slow cooling section. While satisfying the heat preservation requirements, the material discharge cooling requirements are met.

[0106] The furnace chamber of the sintering section is made of corundum muffle bricks, and the top of the furnace chamber is constructed in an arch shape using wedge-shaped bricks. The backing bricks of the corundum muffle bricks are double-layered alumina hollow sphere bricks, the outer layer of which is a furnace wall made of mullite insulating bricks, and the outer layer of the furnace wall is an insulation layer composed of calcium silicate boards.

[0107] Exemplarily, in step S2, the pre-burning section is provided with 3 heating zones, the rated heating temperature is 1100±10℃, the heating mode of the pre-burning section is left-right symmetrical heating, and the rated heating power arranged for each heating zone is 18kW-20kW; the sintering section is provided with 4 heating zones, the rated heating temperature is 1755±5℃, the heating mode of the sintering section is left-right symmetrical heating, and the rated heating power arranged for each heating zone is 45kW-50kW.

[0108] The process system is shown in Table 1 as follows:

[0109] Table 1 Process time of each temperature zone of the high-temperature continuous push-boat sintering furnace

[0110] Temperature / °C Room temperature ~ 500 500～900 900～1750 1750 1750～1760 Time / h 3.5～4 4～5 1.5～2 7～8 9～10

[0111] Uniformity requirement: ±10℃ (1100℃), ±5℃ (1755℃).

[0112] The UO2 pellet sintering process curve is shown in Figure 2 . Among them, in the molybdenum boat assembly, the molybdenum boat is installed on the molybdenum bottom plate through positioning pins. According to the process time, the size of the molybdenum boat and the production capacity, 35 boat positions are determined in the high-temperature continuous push-boat sintering furnace.

[0113] The high-temperature continuous push-boat sintering furnace is mainly composed of a replacement bin (containing a feeding replacement cabin and a discharging replacement cabin), a feeding assembly, a sintering furnace body, a cooling jacket, a discharging assembly, a gas supply system, a tail gas discharge system, a boat bottom plate automatic loading and unloading rotary system, an electrical control and data processing system, etc. The sintering furnace is composed of a furnace shell, a furnace cover assembly, a furnace lining, a heating system, a temperature measuring element, etc. The furnace shell is designed with multiple nozzles and small valves for protective gas flushing and replacement during temperature rising and furnace drying.

[0114] The furnace shell is the supporting structure of the furnace lining, and is made of high-quality carbon steel material by airtight welding. The thickness of the carbon steel reaches 5mm, so that the shell has good heat resistance, corrosion resistance and deformation resistance. The furnace shell is integrally formed and welded in a fully sealed manner. All the welds are uniform and smooth without defects such as pores, slag inclusion and incomplete fusion. All the welds are inspected to prevent hidden defects. The structure design should ensure that there is no stress concentration at the joint. The inside of the furnace shell is designed with a frame rib sandwiched layer for heat preservation, which is filled with heat preservation materials and wrapped with mesh steel plate. This design ensures that the surface temperature of the equipment is around 100℃, reduces the thickness of the furnace lining, and increases the strength of the furnace shell and the support of the side wall refractory bricks. The load-bearing of the refractory bricks helps to reduce the deformation of the refractory bricks. Double-layer hollow sphere alumina refractory bricks are used in the design. The refractoriness of the hollow sphere bricks can reach 1800℃, the density is 1.58 / cm 3 , the porosity is about 50%, although the strength is not as good as that of refractory bricks, but it is much better than that of hot bricks, and the heat insulation effect is better than that of corundum bricks, which is conducive to ensuring the safety of the structure.

[0115] The determination of surface temperature has a great influence on furnace lining design and thermal calculation. If the surface temperature is too low, it will inevitably increase the thickness of the furnace lining, increase the manufacturing cost and the occupied floor area, but reduce the power. Taking the side as an example, the calculation method of temperature and gradient is explained. The surface temperature of the outer side is designed to be 100℃. If the power is too low, it is not conducive to the quality of the sintering of the pellets. Maintaining a surface temperature of 100℃ is optimal for the structure of the furnace lining and the power and product quality. The side bricks use a layer of corundum refractory bricks for fire blocking, three layers of alumina hollow ball bricks for heat insulation and fire blocking, three layers of lightweight mullite heat insulation bricks for heat insulation, and two layers of calcium silicate boards for heat preservation. The selection of each layer of bricks is unified, from the inside to the outside, the refractoriness is from high to low, and the heat preservation effect is from low to high.

[0116] Due to the particularity of the product and the atmosphere of the furnace body, the refractory bricks used for the sintering of nuclear fuel pellets are basically made of corundum. Corundum refractory material refers to refractory material with a-Al2O3(corundum) as the main phase. It has the advantages of high temperature resistance, high hardness, high strength, oxidation resistance, corrosion resistance, good insulation and air tightness, etc. The disadvantages are large linear expansion coefficient, large elastic modulus and poor thermal shock resistance. All previous sintering furnaces are imported, and the refractory bricks for the furnace are also imported from abroad. The imported refractory bricks are manufactured by equipment manufacturers and professional refractory manufacturers. One of the main features of this push boat furnace is that it uses domestic refractory materials throughout the furnace. This is the first time it is used in a domestic 1800℃ high temperature sintering furnace. The production and use characteristics of domestic corundum refractory materials are studied. Corundum is the basic raw material for refractory bricks, mainly in two crystal forms, alumina and high-alumina vanadium clay. To become a corundum material, it must undergo electrosol and high temperature sintering to convert all the crystal forms and eliminate impurities, thereby improving the refractory performance of the corundum material. The performance of corundum powder has a great influence on the performance of refractory products.

[0117] The furnace cover assembly is composed of a furnace cover shell, insulation material, and lifting device mounting plate. The furnace cover shell is designed in sections, with the main body being a whole steel plate, welded for air tightness. All welds are uniform and smooth, without defects such as air holes, slag inclusions, and incomplete fusion. The upper part is welded with flat steel as a reinforcing rib. This ensures that the furnace body has sufficient strength and rigidity, and guarantees good air tightness in the furnace.

[0118] Each section of the furnace cover can be automatically lifted by the automatic lifting device, which meets the demand for flexible maintenance in the case of convenient maintenance. The furnace cover lifting device is mainly used for the automatic lifting of the furnace cover, and is composed of a lifting frame, a pulley block, a limiting device, a lifting tool, a mechanical safety lock, and a motor. The lifting device is designed in sections, corresponding to different parts of the furnace cover, and can facilitate the individual maintenance of each section of the furnace lining.

[0119] The furnace lining, also known as the masonry, is a hearth built with refractory materials, heat insulating materials. For pushboat furnace, its role has two points: one is to form the channel of pushboat; two is that the hearth can withstand high temperature heat load in the process of heating, resist chemical corrosion, reduce heat loss and have high strength, protect the heat exchange in the furnace. It is mainly divided into five temperature sections: import section, pre-burning section, transition section, sintering section and slow cooling section, which are separated by four partition beams. Due to different functions, the structure of each section is different, among which the most important are the pre-burning section and the sintering section. The sintering temperature of the core block is as high as 1750-1760℃, and the furnace lining requires good high-temperature resistance and stability. Since the sintered product is UO2core block, it has high requirements for impurities, and it is in a hydrogen environment atmosphere, so corundum is the best choice for the refractory material of the furnace lining. The existing refractory materials used for UO2core block sintering are basically the same.

[0120] There are two sets of replacement bins respectively arranged at the inlet and outlet, connected with the feeding assembly and the discharging assembly. Each set of replacement bin contains a replacement bin body, a sealing valve, a vacuum purging system, a gas filling connector, an electric pushing mechanism, an inlet (outlet) inner door, an inlet (outlet) outer door, an outside temporary storage track station, an observation window, a boat bottom plate in place detection device and other components. The bin body is provided with a pressure sensor and an oxygen analyzer.

[0121] The replacement bin body is composed of a bin chamber shell, a sliding rail, an air extraction and gas filling curtain box, and can accommodate one set of molybdenum boat and boat bottom plate combination. The gas replacement bin chamber is entirely made of high-quality carbon steel, adopts a single-layer structure, and the outside is welded with a shaped steel as a reinforcing rib. The shell bottom is fixed with a sliding rail, and the push plate moves on the sliding rail.

[0122] The feeding assembly is the part connecting the inlet replacement bin and the sintering furnace, which is composed of an inlet bin, a main pushing mechanism, a rack, a in-place sensor, an access hole, etc.

[0123] The inlet bin is made of high-quality carbon steel material. The bin body adopts a single-layer structure, and the outside is welded with a shaped steel as a reinforcing rib.

[0124] The shell bottom is fixed with a sliding rail, and the push plate moves on the sliding rail. The main pushing mechanism is controlled and driven by an electric servo cylinder, installed at one end of the inlet bin furnace tube, composed of a pushing mechanism, a push head plate, a guide sliding rail, etc.

[0125] Main push force calculation.

[0126] (1) Main push function and main technical parameters:

[0127] The main push is arranged at the feeding end of the furnace to push the tray into the furnace for heating. The travel limit and overload protection device are provided on it to ensure safety in production.

[0128] The main parameters of the molybdenum lanthanum alloy boat and the molybdenum lanthanum alloy push plate include:

[0129] Full load: 45kg; total length of push: 11960mm; main push forward speed: 0-15mm / min, uniform speed push boat; main push backward speed: ≥340mm / min, deceleration before returning to position, avoiding fast return.

[0130] Main push displacement: 400-600mm; take 600mm.

[0131] (2) Main push force calculation.

[0132] F=G×g×e×K1×n

[0133] In the formula, F is the main push force, unit N; G is the total mass of the push plate + boat + material, the total weight of a single push plate + boat + material is 75kg, and the total weight of 35 boats in the furnace is 2625kg. e is the sliding friction factor, e=0.8-1 when the temperature is greater than 800℃, and e=0.95. g is the acceleration of gravity, 9.8m / s. K1 is the influence factor considering track unevenness, thermal deformation and other factors, K1=1.1. n is the safety factor, n=1.8.

[0134] (3) Selection of electric cylinder.

[0135] According to F=50KN, the selection is shown in Table 2;

[0136] Table 2 Electric cylinder configuration scheme

[0137]

[0138] (4) According to the selection calculation, it can be obtained that the main push force of TJEN135S830BT10BM2C1 type servo electric cylinder has 80% of the surplus amount, which fully meets the requirements.

[0139] In step S2, a modified average power distribution method is used to smoothly distribute the heating power of each heating zone, including:

[0140] A heating power network framework based on generative adversarial imitation learning is constructed, and the specific parameters are shown in Tables 3 and 4;

[0141] Table 3 Network parameter settings of heating power network framework

[0142]

[0143] Table 4 Parameter settings of main program of heating power network

[0144]

[0145] The expert data is used to train the discrimination module and the generation module, and the discrimination judgment is confused, and the discrimination module is used to distinguish the data distribution generated by the generation module; the power is initialized as P0, and the sampling trajectory is The power data sampling trajectory generated by the generation module is The sampling trajectory of the real power in a certain heating area is

[0146] The power distribution mode function is updated by gradient descent, and the discrimination module is updated by gradient descent, so as to distinguish the weight distribution of real data and generated data, and the expression is:

[0147]

[0148] In the formula, The power distribution mode function value updated by gradient descent is f, the number of heating areas is z (i) The real data of a certain heating area node is H(k (i) The weight of the generated data is The weight judgment value of the real power in the sampling trajectory of a certain heating area is U(H(k (i) The weight judgment value of the generated data is The weight judgment value in the power data sampling trajectory generated by the generation module is U(), and the weight judgment is U(z (i) The weight judgment of the real data by the discrimination module is

[0149] The loss function of the generated adversarial network is constructed, and the reward function is constructed by the discrimination module, and the generation module is trained, and the expression is:

[0150]

[0151] In the formula, The loss function value of the generated adversarial network updated by gradient descent is

[0152] When the discrimination module cannot distinguish the data generated by the generation module from the real data, the generation module and the discrimination module reach Nash equilibrium, and then the generation module successfully matches the expert power distribution mode; the specific parameters are shown in Table 5;

[0153] Table 5 Parameter settings based on improved average power distribution mode optimization method

[0154] Parameter Value / class Running equipment GPU Maximum iteration number 500 Learning rate 9.9234*10 -6 ]]>

[0155] The improved average power distribution mode optimization method is used to optimize the parameters of the generation module network, and the power distribution uniform sample based on the improved average power distribution mode optimization method is as follows:

[0156]

[0157] In the formula, F CLIP (ω) represents the optimized power distribution uniformity value. For the average power distribution optimization method, min() is the minimum function value, e t (ω) represents the weighting of the heating power allocation for a certain heating zone. ρ is the estimated value at a certain moment, clip() is the objective function for modifying the heating power of the replacement heating zone by the cutoff threshold, and ρ is the cutoff threshold.

[0158] The parameters of the generator module network are optimized based on the improved average power allocation method. The specific parameters are shown in Table 6.

[0159] Table 6 Network parameter settings for the identification module

[0160]

[0161]

[0162] The specific steps include:

[0163] (1) Estimation of the dominance of generalized average allocation

[0164] Compare the power obtained by performing heating action J under a steady current P with the power allocation method P(T|J) to evaluate the suitability and deviation of heating action J, and update the weight M(J|T) of heating action J in the direction that the power allocation method is greater than zero. The generalized average allocation advantage estimate is as follows:

[0165]

[0166] In the formula, Let (μ, η) be the estimate of the generalized average distribution advantage, where ∞ represents infinity. n The values ​​are constrained by the heating coefficient and discount rate for n iterations. Let W(T) be the heating power state value function at the current moment. t+n+1 (e) is a function of the current value in the heating state at the next moment. t+n Let W(T) be the power function at the current moment. t+n ) represents the steady current value function at the current moment, and η represents the discount rate of the value function at the previous moment;

[0167] (2) Stable current - heating action weighting ratio e t (ω), expressed as:

[0168]

[0169] In the formula, P ω(J t |T t ) is the logarithmic weight of a batch of heating action sets performed by the new power distribution mode, P old (J t |T t ) is the average heating action logarithmic weight of the old power distribution mode.

[0170] From the above embodiments, compared with the traditional behavior feature extraction, the method establishes an accurate power distribution mode mathematical model; the power distribution mode learning model based on the application has good robustness and accuracy.

[0171] Exemplarily, in step S3, the self-adjustment of the PID parameters by using the improved fuzzy control rule includes:

[0172] (1) Inference of the temperature change amount: the temperature change amount is divided into less, normal, and high;

[0173] The control rule of the temperature change amount is: ① if the furnace temperature field distribution change amount is small, the temperature change amount is less; ② if the furnace temperature field distribution change amount is moderate, the temperature change amount is normal; and ③ if the furnace temperature field distribution change amount is more, the temperature change amount is high;

[0174] (2) Inference of the heating temperature of the heating body: the heating temperature is divided into slightly low, normal, and high;

[0175] The control rule of the heating temperature of the heating body is: ① if the furnace temperature field distribution change amount is small, the heating temperature is slightly low; ② if the furnace temperature field distribution change amount is moderate, the heating temperature is normal; and ③ if the furnace temperature field distribution change amount is more, the heating temperature change amount is high;

[0176] (3) Inference of the PID parameter adjustment: the PID parameter adjustment is divided into short, short, moderate, long, and long; the rules of the five fuzzy modes are: ① if the furnace temperature field distribution change amount is small and the temperature change degree is low, the PID parameter adjustment is short; ② if the furnace temperature field distribution change amount is small and the temperature change degree is moderate, the PID parameter adjustment is short; ③ if the furnace temperature field distribution change amount is small and the temperature change degree is normal, the PID parameter adjustment is moderate; ④ if the furnace temperature field distribution change amount is moderate and the temperature change degree is low, the PID parameter adjustment is short; ⑤ if the furnace temperature field distribution change amount is moderate and the temperature change degree is moderate, the PID parameter adjustment is moderate; ⑥ if the furnace temperature field distribution change amount is moderate and the temperature change degree is high, the PID parameter adjustment is long; ⑦ if the furnace temperature field distribution change amount is less and the temperature change degree is low, the PID parameter adjustment is moderate; ⑧ if the furnace temperature field distribution change amount is more and the temperature change degree is moderate, the PID parameter adjustment is long; and ⑨ if the furnace temperature field distribution change amount is more and the temperature change degree is high, the PID parameter adjustment is long;

[0177] The fuzzy inference result is defuzzified, the inference result is converted into an accurate value, the maximum entropy weight value method is used in the control system, the maximum entropy weight value method does not consider the shape of the output entropy weight value function, only considers the output value at the maximum entropy weight value, the maximum entropy weight value defuzzification method selects the element with the maximum entropy weight value in the fuzzy quantity of the inference result as the output value, and the expression is:

[0178]

[0179] In the formula, Γ0 is the output value of the element with the maximum entropy weight value in the fuzzy quantity of the inference result, max() is a maximum function, It is a calculation mode of the maximum entropy weight value defuzzification, Γ is the output quantity of the maximum entropy weight value defuzzification, and S is an output universe.

[0180] If the output value corresponding to the maximum entropy weight value is more than one in the output universe S, the average value of the output with the maximum entropy weight value is taken, and the expression is:

[0181]

[0182] In the formula, I is the total number of the outputs with the maximum same entropy weight value, Γ i It is the value of the output of a certain entropy weight value.

[0183] It can be known from the above embodiment that the improved fuzzy control rule is used for self-adjusting PID parameters, and the control precision is high.

[0184] In another example, in step S1, in the water cooling section, the electrical control system with the PLC as the main control unit also monitors and controls the tail gas discharge system, the electrical control system with the PLC as the main control unit collects the process parameters of the gas supply system in each temperature section by using a plurality of different sensors, the process parameters include gas pipeline pressure, water pipeline pressure, furnace pressure, furnace temperature, dew point and water temperature, and the signals generated by the collected process parameters are transmitted to the PLC and the industrial computer through a line for monitoring, are uniformly controlled in the integrated industrial configuration system, and are used for omnibearing monitoring of the operation and stable current of the high-temperature continuous push boat sintering furnace, and automatic setting and monitoring of the process parameters are completed.

[0185] The gas supply system comprises a hydrogen gas supply system, a nitrogen gas supply system, a natural gas supply system and a compressed air supply system.

[0186] As shown in the figure, the embodiment of the present application provides a high-temperature continuous push boat sintering furnace temperature uniformity control system, which comprises: Figure 4

[0187] ​The temperature section division and insulation material selection module 1 is used for dividing the sintering furnace into a pre-sintering section, a sintering section, a slow cooling section and a water cooling section according to process requirements, arranging heating bodies in the pre-sintering section and the sintering section, and selecting different insulation materials according to insulation parameters of different temperature sections;

[0188] The pre-sintering zone furnace structure comprises a pre-sintering section, wherein the pre-sintering section comprises a hoisting furnace cover, a plurality of layers of mullite heat insulation bricks and a silicon-calcium plate are arranged on the left and right sides of the hoisting furnace cover, an electrode penetrates through the silicon-calcium plate and the plurality of layers of mullite heat insulation bricks to connect a heating body, a corundum brick is further arranged outside the heating body and located between the innermost layer of mullite heat insulation bricks and the heating body, the heating body is located below the hoisting furnace cover, a corundum mullite muffle is arranged inside the heating body, a molybdenum boat is arranged inside the corundum mullite muffle, and a slide plate is arranged below the molybdenum boat. A plurality of layers of mullite heat insulation bricks are further arranged below the corundum mullite muffle.

[0189] The highest design temperature of the pre-sintering zone is 1100±10℃, and the actual use temperature is generally not higher than 900℃. The pre-sintering zone is divided into three temperature control zones, and the temperature gradient is formed from 500-900℃. The pre-sintering zone has the functions of preheating the sintering product and removing additives such as AKL and ammonium oxalate in the briquette under certain temperature conditions. A large amount of volatile substances are generated in the pre-sintering zone, which is not conducive to the cleaning of the furnace and the service life of the heating element in the later maintenance operation. Therefore, the closed muffle design is adopted in the pre-sintering section, the pre-sintering section is divided into several sections, the muffle is divided into three parts, namely the upper muffle, the left muffle and the right muffle, and the muffle can be removed during maintenance to facilitate the cleaning of volatile substances. The muffle is made of corundum mullite and has good thermal conductivity and thermal shock resistance.

[0190] The average power distribution module 2 is used for dividing the pre-sintering section and the sintering section into multiple heating zones and smoothly distributing the heating power of each heating zone by using an improved average power distribution method, so that the temperature of the pre-sintering section and the sintering section is smoothly transitioned.

[0191] The temperature adjustment module 3 is used for monitoring the temperature change data information of the pre-sintering section and the sintering section, adjusting the PID parameters by using an improved fuzzy control rule, and realizing the smooth distribution of the furnace temperature in the furnace.

[0192] The electrical control system 4 is used for taking the PLC as a main control unit, collecting the process parameters of the gas supply system of each temperature section by using multiple different sensors, transmitting the signals generated by the collected process parameters to the PLC and the industrial computer through a line for monitoring, uniformly controlling in the integrated industrial configuration system, comprehensively monitoring the operation and stable current of the high-temperature continuous push boat sintering furnace, and completing the automatic setting and monitoring of the process parameters.

[0193] The electrical control system 4 takes PLC as the main control unit, and also monitors and controls the tail gas emission system.

[0194] Thermal calculation process: set the insulation layer as a stable flat layer heat conduction mode, first need to obtain the comprehensive heat transfer coefficient a2 of the furnace wall outer wall to the air, its calculation formula is:

[0195]

[0196] In the formula, The furnace wall outer wall temperature is ℃ (need to preset a value); 4.65 represents the ambient temperature ℃; K represents a constant, the value is different in different directions, respectively: side k=2.56; top k=3.26; bottom k=2.1;

[0197] Secondly, calculate the heat flow Q, take 1m 2 The calculation unit is calculated, and the calculation formula is:

[0198]

[0199] In the formula, t1 is the furnace temperature, t f2 The environmental temperature is n, the number of layers of insulation material, δ is the thickness of the insulation material, λ is the thermal conductivity; The specific parameters are shown in Table 7;

[0200] According to the above formula, the interface temperature of each insulation layer can be obtained, and the expression is:

[0201]

[0202] Table 7 Furnace side temperature checking table

[0203]

[0204]

[0205] According to the calculation, the furnace shell surface temperature is 154℃, and the surface temperature after increasing the hanging plate is less than 50℃.

[0206] The heating system is composed of pre-burning section heating elements, sintering section heating elements, and temperature control system. The pre-burning section heating zone has 3 zones, and the sintering section heating zone has 4 zones, both of which adopt the heating method of suspending heaters on the left and right side walls. Each heating zone forms an independent temperature control loop, and each heating zone adopts SCR power controller for heating power control. Each heating zone adopts an independent temperature controller for PID control of temperature. The heating elements are made of multiple strands of molybdenum wire woven together and fixed on the left and right side bricks of the furnace by molybdenum hooks. The two loops are independently temperature controlled. The heater is hung on the left and right sides of the furnace, and the electrode lead rod is used to lead out of the furnace body. The wiring is connected to the sintering electrode rod by binding wire. The electrode rod is sealed with the furnace wall, effectively isolating the atmosphere inside and outside the furnace. A water-cooled clamp plate is installed on the electrode lead flange to ensure the reliability of the sealing element and improve the service life of the sealing element. The cold clamp plate has high heat exchange efficiency and is double-layer insulated with the furnace shell, ensuring that the cooling water does not enter the furnace in the event of a water leakage accident. At the same time, it can effectively reduce the temperature of the electrode rod and prevent the wiring from being oxidized for a long time, causing poor electrical contact and affecting production. This structure not only ensures the cooling efficiency but also increases the safety.

[0207] The heating furnace is provided with 20 thermocouples in total. There are 6 thermocouples in the pre-burning section, 2 in the transition section, 10 in the sintering section, 2 in the slow cooling section, and 2 in the rapid cooling section. The pre-burning section thermocouples adopt K-type double thermocouples, and the sintering section adopts W-type tungsten-rhenium 3-25 double thermocouples. Each heating zone is equipped with one set of double thermocouples (two sets of thermocouples are installed in high-temperature zones 1 and 4), one for temperature control and the other for temperature monitoring. Each set has one safety reserve. The transition section and the slow cooling section both use W-type tungsten-rhenium 3-25 double thermocouples for temperature monitoring. Each temperature zone is equipped with two sets of double thermocouples, each for temperature monitoring, and each set has one safety reserve. The rapid cooling section thermocouples use K-type double thermocouples for temperature monitoring. Each temperature zone is equipped with two sets of double thermocouples, each for temperature monitoring, and each set has one safety reserve. Double thermocouples can effectively reduce the impact of thermocouple failures on equipment production.

[0208] The furnace tail discharge assembly is the part connecting the sintering furnace and the discharge replacement bin, which is composed of forced water cooling section, furnace tail outlet bin, explosion relief device, discharge cylinder, emergency door, emergency maintenance platform, etc.

[0209] The forced water cooling section is composed of a steel muffle and a water cooling assembly. The steel muffle adopts a single-layer structure, and the inner bottom is fixed with a slide rail. The push plate moves on the slide rail. The water cooling assembly is tightly pressed to the outer surface of the muffle by bolts, and high-temperature heat-conducting silicone grease is filled between them. The water cooling assembly is composed of four heat exchange zones of water jackets, which are fixed on the four faces of the muffle. The water jackets are made of stainless steel by welding. The water jackets have the highest contact area with the muffle, high thermal conductivity, which not only avoids the risk of water seeping into the muffle but also improves the cooling efficiency.

[0210] The furnace tail discharge bin is composed of an outlet bin body, an inspection door, a cam height cutting mechanism, a photoelectric detection device, an observation window and the like. The bin body adopts a single-layer structure, and a profile steel is welded outside as a reinforcing rib. A slide rail is fixed at the bottom of the shell, and a push plate moves on the slide rail.

Claims

1. A method for regulating the temperature uniformity of a high-temperature continuous pusher boat sintering furnace, characterized by, The method comprises the following steps: S1: According to the process requirements, the sintering furnace is divided into a pre-sintering section, a sintering section, a slow cooling section and a water cooling section, and heating bodies are arranged in the pre-sintering section and the sintering section; different heat preservation materials are selected for masonry according to the heat preservation parameters of different temperature sections; S2: The pre-sintering section and the sintering section are divided into multiple heating zones, and an improved average power distribution method is used to smoothly distribute the heating power of each heating zone, so that the temperature of the pre-sintering section and the sintering section is smoothly transitioned; S3: The temperature change data information of the pre-sintering section and the sintering section is monitored, and an improved fuzzy control rule is used for PID parameter self-adjustment to realize the smooth distribution of the furnace temperature in the hearth.

2. The method for regulating the temperature uniformity of a high-temperature continuous pusher boat sintering furnace according to claim 1, characterized in that: In the step S1, the pre-sintering section and the sintering section are connected through a transition section, no heating body is arranged in the slow cooling section and the water cooling section, and the thickness of the furnace lining of the water cooling section is less than that of the slow cooling section.

3. The method for regulating the temperature uniformity of a high-temperature continuous pusher boat sintering furnace according to claim 2, characterized in that: The hearth of the sintering section is a corundum muffle masonry with corundum bricks, the top of the hearth is wedge-shaped and arched, the back-up bricks of the corundum muffle are double-layered alumina hollow ball bricks, the outer layer of the double-layered alumina hollow ball bricks is a furnace wall masonry with mullite insulating bricks, and the outer layer of the furnace wall is an insulating layer composed of silicon-calcium plates.

4. The method for regulating the temperature uniformity of a high-temperature continuous pusher boat sintering furnace according to claim 1, characterized in that: In the step S2, the pre-sintering section is provided with three heating zones, the heating temperature is 1100±10℃, the pre-sintering section is heated in a left-right symmetrical manner, and the heating power arranged in each heating zone is 18kW-20kW; the sintering section is provided with four heating zones, the heating temperature is 1755±5℃, the sintering section is heated in a left-right symmetrical manner, and the heating power arranged in each heating zone is 45kW-50kW.

5. The method for regulating the temperature uniformity of a high-temperature continuous pusher boat sintering furnace according to claim 1, characterized in that: In the step S2, the improved average power distribution method is used to smooth the distribution of the heating power of each heating zone, including: constructing a heating power network framework based on generative adversarial imitation learning, training a discrimination module and a generation module with expert data, confusing the judgment of the discrimination, and using the discrimination module to distinguish the data distribution generated by the generation module; initializing the power as P0, and the sampling trajectory as the power data sampling trajectory generated for the generation module, the sampling trajectory of the real power in a certain heating zone; The expression is: In the formula, is the function value of the power allocation mode of the gradient descent update, f is the number of heating zones, z (i) is the real data of a heating zone node, H(k (i) ) is the weight of the generated data, is the weight determination value of the real power in the sampling trajectory of a heating zone, U(H(k (i) )) is the weight determination value of the generated data, is the weight determination value in the power data sampling trajectory generated by the generation module, U() is the weight determination, U(z (i) ) is the weight determination of the real data by the discrimination module; The loss function of the generative adversarial network is constructed, and the reward function is constructed with the discriminant module to train the generative module, and the expression is: In the formula, is the loss function value of the gradient descent update of the generative adversarial network. When the discriminant module cannot distinguish the data generated by the generative module from the real data, the generative module and the discriminant module reach Nash equilibrium, and then the generative module successfully matches the expert power distribution mode; The improved average power distribution optimization method is used for parameter optimization of the generative module network, and the power distribution uniform sample based on the improved average power distribution optimization method is as follows: In the formula, F CLIP (ω) is the optimized power distribution uniform value, is the average power distribution optimization method, min() is the minimum function value, e t (ω) is the heating power distribution weight of a certain heating area, is the estimated value at a certain moment, clip() is the truncated threshold modification alternative heating power target function, and p is the truncated threshold.

6. The method for regulating the temperature uniformity of a high-temperature continuous pusher boat sintering furnace according to claim 5, characterized in that: The improved average power distribution optimization method is used for parameter optimization of the generative module network, and the improved average power distribution optimization method is used for parameter optimization of the generative module network, and the improved average power distribution optimization method is used for parameter optimization of the generative module network. (1) Generalized average allocation advantage estimation The power obtained by executing the heating action J under the stable current T and the current power distribution mode P(T|J) are compared to evaluate the suitability and deviation of the heating action J, so that the power distribution mode is updated to the direction in which the average distribution advantage estimation is greater than zero; the generalized average distribution advantage estimation is as follows: wherein is a generalized average distribution advantage estimate value, ∞ is infinity, (μ,η) n is a heating coefficient, a discount rate constraint value for n times of circulation, is a heating power state value function at the current time, W(T t+n+1 ) is a current value function of the heating state at the next time, e t+n is a power function at the current time, W(T t+n ) is a stable current value function at the current time, and η is a discount rate of the value function at the previous time. (2) The stable current-heating operation weight ratio e t (ω), expressed by the expression: where P ω (J t |T t ) is the log weight of a batch of heating action sets performed by the new power allocation method, P old (J t |T t ) is the log weight of an average heating action performed by the old power allocation method.

7. The method for regulating the temperature uniformity of a high-temperature continuous pusher boat sintering furnace according to claim 1, characterized in that: In the step S3, the improved fuzzy control rule is used for PID parameter self-adjustment, which comprises the following steps: the fuzzy inference result is de-fuzzified, the inference result is converted into an accurate value, the maximum entropy weight value method is adopted, the maximum entropy weight value de-fuzzification method selects the element with the maximum entropy weight value in the fuzzy quantity of the inference result as the output value, and the expression is: Γ0= max(θ Γ (Γ)), Γ ∈ S In the formula, Γ0 is the output value of the element with the maximum entropy weight value in the fuzzy quantity of the inference result, max() is the maximum function, θ Γ () is the calculation method of the maximum entropy weight value inverse fuzzification, Γ is the output quantity of the maximum entropy weight value inverse fuzzification, and S is the output domain. If the output value corresponding to the maximum entropy weight value is more than one in the output universe S, the average value of the output with the maximum entropy weight value is taken, and the expression is: where I is the total number of the maximum identical entropy weight outputs, Γ i is the value of the certain entropy weight output.

8. A system for regulating the temperature uniformity of a high-temperature continuous pusher boat sintering furnace, which is realized by the method for regulating the temperature uniformity of a high-temperature continuous pusher boat sintering furnace according to any one of claims 1-7, characterized in that: The system comprises: The temperature section division and insulation material selection module is used for dividing the sintering furnace into pre-sintering section, sintering section, slow cooling section and water cooling section according to process requirements, arranging heating bodies in the pre-sintering section and the sintering section, and selecting different insulation materials according to insulation parameters of different temperature sections; The average power distribution module is used for dividing the pre-sintering section and the sintering section into multiple heating zones, and using an improved average power distribution method to smoothly distribute the heating power of each heating zone, so that the temperature of the pre-sintering section and the sintering section is smoothly transitioned; The temperature adjustment module is used for monitoring temperature change data information of the pre-sintering section and the sintering section, using an improved fuzzy control rule to adjust PID parameters automatically, and realizing smooth distribution of the furnace temperature in the hearth.

9. The furnace temperature uniformity regulating system of a high-temperature continuous pusher boat sintering furnace according to claim 8, characterized in that: The system further comprises an electrical control system, which uses PLC as a main control unit, collects process parameters of a gas supply system of each temperature section by using multiple different sensors, transmits signals generated by the collected process parameters to PLC and an industrial computer through a circuit for monitoring, comprehensively monitors operation and stable current of the high-temperature continuous push boat sintering furnace, and completes automatic setting and monitoring of process parameters. The electrical control system uses PLC as a main control unit to monitor and control a tail gas discharge system.

Images