Intelligent control method and system for vacuum annealing of die steel

Abstract

The present invention provides a mold steel vacuum annealing intelligent control method and system, which relates to the field of intelligent control technology, including dividing the target mold steel and the interior of the furnace where the target mold steel is located into a plurality of finite element units, determining the radiation intensity according to the preset radiation diffusion coefficient and the temperature of any point in the finite element unit in the target mold steel, combining the radiation model and the radiation boundary condition; iteratively coupling the radiation heat transfer coefficient matrix and the radiation intensity, and calculating the three-dimensional transient temperature field distribution inside the target mold steel; taking the temperature deviation as input, performing fuzzy reasoning, and obtaining the PID controller parameters after adaptive adjustment; substituting the PID controller parameters after adaptive adjustment into the PID control algorithm, obtaining the control output of the composite controller, and applying the control output to the heater of the vacuum annealing furnace of the target mold steel, so as to realize the control of the annealing temperature. The present invention can realize the intelligent control of the vacuum annealing of the mold steel, and improve the accuracy of the annealing temperature control.

Description

Technical Field

[0001] The invention relates to intelligent control technology, and in particular to an intelligent control method and system for vacuum annealing of mold steel. Background Art

[0002] Mold steel is an important material for manufacturing precision molds, and its performance directly affects the quality and service life of the mold. In order to obtain excellent comprehensive performance, mold steel usually needs to be vacuum annealed. Vacuum annealing is a heat treatment process performed in a vacuum environment. By controlling parameters such as heating temperature, holding time and cooling rate, it can effectively eliminate the residual stress inside the mold steel, stabilize its size, improve the plasticity and toughness of the material, and improve the cutting performance.

[0003] Traditional vacuum annealing control mainly relies on the operator's experience and skills, and has the following problems:

[0004] Low control accuracy: It is difficult to achieve precise adjustment of parameters such as temperature and time through manual control, which can easily lead to fluctuations and instability in annealing quality.

[0005] High energy consumption: Non-optimal control strategies may result in excessively long heating time and excessively high temperatures, resulting in unnecessary energy waste.

[0006] High labor intensity: Operators need to monitor the annealing process for a long time and make manual adjustments based on experience. The work intensity is high and it is easy to get tired and make mistakes.

[0007] Lack of intelligence: Traditional control methods are difficult to adapt to the annealing requirements of different steel grades and workpieces of different specifications, and lack flexibility and adaptability. Summary of the invention

[0008] The embodiments of the present invention provide a mold steel vacuum annealing intelligent control method and system, which can solve the problems in the prior art.

[0009] A first aspect of an embodiment of the present invention provides an intelligent control method for vacuum annealing of mold steel, comprising:

[0010] The target mold steel and the interior of the furnace body where the target mold steel is located are divided into a plurality of finite element units, and a radiation heat transfer coefficient matrix between the finite element units inside the furnace body is calculated using a ray tracing algorithm; the radiation intensity is determined based on a preset radiation diffusion coefficient and the temperature of any point in the finite element unit in the target mold steel, in combination with a radiation model and a radiation boundary condition;

[0011] Iteratively coupling the radiation heat transfer coefficient matrix and the radiation intensity, and obtaining the three-dimensional transient temperature field distribution inside the target mold steel through interpolation calculation;

[0012] According to the three-dimensional transient temperature field distribution and the target temperature setting value, the temperature deviation is calculated; with the temperature deviation as input, fuzzy reasoning is performed through a preset fuzzy control rule library to obtain adaptively adjusted PID controller parameters; the adaptively adjusted PID controller parameters are substituted into the PID control algorithm to calculate the control output of the composite controller, and the control output is applied to the heater of the target mold steel vacuum annealing furnace to achieve control of the annealing temperature.

[0013] Optionally, determining the radiation intensity according to a preset radiation diffusion coefficient and the temperature of any point in a finite element unit in the target mold steel in combination with a radiation model and a radiation boundary condition includes:

[0014] ;

[0015] in, represents the diffusion coefficient, κ represents the absorption coefficient, σ represents the Stefan-Boltzmann constant, T Indicates the material temperature, G Indicates the radiation intensity;

[0016] Radiation boundary conditions include:

[0017] ;

[0018] ε represents the emissivity of the material surface, T w represents the material temperature at the boundary of the finite element unit in the target mold steel, G w represents the radiation intensity at the boundary of the finite element unit in the target mold steel, n Represents the number of finite element elements.

[0019] Optionally, the method further comprises calculating the radiation heat transfer coefficient according to a ray tracing algorithm:

[0020] Randomly emitting light inside the furnace where the target mold steel is located, the starting point, direction and energy of each light ray are randomly generated according to the temperature and emissivity of any point in the finite element unit of the target mold steel;

[0021] For each ray, the incident angle and reflection angle of the ray are calculated using the intersection of the ray and the finite element unit, and the absorptivity and reflectivity of the ray are calculated based on the material properties of the finite element unit;

[0022] Track the propagation path of the light inside the furnace until the light is completely absorbed or escapes from the furnace;

[0023] The light energy absorbed by each finite element unit is counted, and the radiation heat transfer coefficient matrix between finite element units is calculated.

[0024] In an optional embodiment,

[0025] The calculation of the radiation heat transfer coefficient matrix includes:

[0026] ;

[0027] in, Indicates the number of mold steel parts inside the furnace. i Finite element units to j The radiation heat transfer coefficient of the finite element unit is N Indicates the number of rays, θ ik , θ jk Respectively represent k The light and i and j The angle between the normal vectors of the finite element, r ijk Indicates k The ray is in i and j The distance between the finite element units, V k Indicates k The reflectivity of the light, δ k Indicates k The absorption rate of a ray of light.

[0028] Optionally, taking the temperature deviation as input, fuzzy reasoning is performed through a preset fuzzy control rule base to obtain adaptively adjusted PID controller parameters including:

[0029] The temperature deviation is fuzzified by a triangular function with a variable width to obtain a temperature fuzzy value;

[0030] Through the preset fuzzy control rule base, the rule matching degree between the temperature fuzzy value and each fuzzy rule in the fuzzy control rule base is determined by the Mamdani reasoning method, and the corresponding matching weight is assigned to the rule matching degree, and the fuzzy rules are weightedly aggregated to obtain the reasoning result;

[0031] An adaptive scaling factor is introduced, and the inference result is defuzzified by using the maximum membership method to obtain adaptively adjusted PID controller parameters.

[0032] Optionally, before the temperature deviation is fuzzified by a triangular function of variable width to obtain a temperature fuzzy value, the method further comprises constructing a triangular function of variable width:

[0033] The historical data of temperature deviation acquired in advance are clustered by fuzzy mean clustering algorithm to obtain multiple cluster centers;

[0034] The left width and right width of each cluster center are calculated according to the following formulas:

[0035] ;

[0036] ;

[0037] in, Indicates the left width, Indicates the right width, v c Indicates c Cluster centers, x h Indicates h Historical data, Indicates h Historical data for c The membership degree of the cluster centers is b represents the fuzzification parameter;

[0038] According to each cluster center, left width and right width, a variable width triangle function is constructed according to the following formula:

[0039] ;

[0040] in, μ c (x h ) Represents the value of a triangular function.

[0041] Optionally, before substituting the adaptively adjusted PID controller parameters into a PID control algorithm to calculate the control output of the composite controller, the method further includes optimizing the controller parameters of the composite controller:

[0042] The controller parameters of the composite controller are optimized by particle swarm optimization algorithm;

[0043] The adaptively adjusted PID controller parameters are used as the particle positions, the particle speeds are defined, and the positions and speeds of a group of particles are randomly initialized to form an initial particle group;

[0044] According to the performance indicators of the control system, a fitness function is selected to evaluate the quality of the controller parameters corresponding to each particle;

[0045] In each iteration, the particle adjusts its speed and position according to its own optimal position and the global optimal position through the speed and position update formula. Inertia weight, acceleration constant and random number are introduced into the speed and position update formula. In the iterative process, the individual optimal solution and the global optimal solution of the particle are continuously updated until the preset termination condition is reached, and the optimized composite controller parameters are obtained.

[0046] A multi-group collaboration mechanism is introduced to divide the initial particle group into multiple sub-groups, and the collaboration and competition between sub-groups are realized through information exchange and migration operations.

[0047] A second aspect of the embodiment of the present application provides a mold steel vacuum annealing intelligent control system, comprising:

[0048] The first unit is used to divide the target mold steel and the interior of the furnace body where the target mold steel is located into a plurality of finite element units, and calculate the radiation heat transfer coefficient matrix between the finite element units inside the furnace body by using a ray tracing algorithm; and determine the radiation intensity according to a preset radiation diffusion coefficient and the temperature of any point in the finite element unit in the target mold steel, in combination with a radiation model and a radiation boundary condition;

[0049] The second unit is used for iteratively coupling the radiation heat transfer coefficient matrix and the radiation intensity, and obtaining the three-dimensional transient temperature field distribution inside the target mold steel through interpolation calculation;

[0050] The third unit is used to calculate the temperature deviation according to the three-dimensional transient temperature field distribution inside the target mold steel and the target temperature setting value; take the temperature deviation as input, perform fuzzy reasoning through a preset fuzzy control rule library, and obtain the adaptively adjusted PID controller parameters; substitute the adaptively adjusted PID controller parameters into the PID control algorithm, calculate the control output of the composite controller, and apply the control output to the heater of the vacuum annealing furnace of the target mold steel to achieve control of the annealing temperature.

[0051] According to a third aspect of an embodiment of the present invention, there is provided an electronic device, including:

[0052] processor;

[0053] a memory for storing processor-executable instructions;

[0054] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0055] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described above is implemented.

[0056] Compared with the existing technology, this application divides the target mold steel and the interior of the furnace into finite element units, uses a ray tracing algorithm to calculate the radiation heat transfer coefficient matrix, and determines the radiation intensity in combination with the radiation model and boundary conditions. Through iterative coupling and interpolation calculations, the three-dimensional transient temperature field distribution inside the target mold steel is accurately obtained, providing accurate data support for subsequent temperature control.

[0057] According to the three-dimensional transient temperature field distribution and the target temperature setting value, the temperature deviation is calculated. With the temperature deviation as input, fuzzy reasoning is performed through the preset fuzzy control rule library to adaptively adjust the PID controller parameters to achieve intelligent and adaptive control of the annealing temperature, which can improve the control accuracy and robustness. The adaptively adjusted PID controller parameters are substituted into the PID control algorithm to calculate the control output of the composite controller, and the control output is applied to the heater of the target mold steel vacuum annealing furnace to achieve precise control of the annealing temperature. The composite control strategy combines the intelligence of fuzzy control and the stability of PID control, which can effectively deal with interference and uncertainty in the annealing process and ensure the stability of annealing quality.

[0058] By precisely controlling the annealing temperature field distribution, the target mold steel is ensured to be heated evenly and the temperature gradient is reasonable during the annealing process, which can effectively eliminate the residual stress inside the material, stabilize the size, improve the plasticity and toughness, and improve the cutting performance, thereby improving the comprehensive performance and service life of the mold steel. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a schematic diagram of the process of the intelligent control method for vacuum annealing of mold steel according to an embodiment of the present invention;

[0060] Figure 2 It is a structural schematic diagram of the intelligent control system for vacuum annealing of mold steel according to an embodiment of the present invention. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0063] Figure 1FIG. 1 is a flow chart of an intelligent control method for vacuum annealing of mold steel according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0064] S101. Divide the target mold steel and the interior of the furnace body where the target mold steel is located into a plurality of finite element units, and calculate the radiation heat transfer coefficient matrix between the finite element units inside the furnace body by using a ray tracing algorithm; determine the radiation intensity according to a preset radiation diffusion coefficient and the temperature of any point in the finite element unit in the target mold steel, in combination with a radiation model and a radiation boundary condition;

[0065] Exemplarily, according to the geometric shape and size of the target mold steel, finite element analysis software is used to perform three-dimensional modeling on it; according to the structure and layout inside the furnace body, the internal space of the furnace body is three-dimensionally modeled; the target mold steel model and the internal space model of the furnace body are assembled to form a complete furnace body-workpiece system model; the furnace body-workpiece system model is meshed to generate a number of finite element units, each of which contains node and boundary surface information.

[0066] For each finite element unit inside the furnace body, the ray tracing algorithm is used to simulate the radiation heat transfer process; starting from the center point of each unit, a number of rays are emitted to the center points of other units, and the direction of the rays follows a certain angle distribution rule; for each ray, its intersection with the boundary surface of other units is calculated, and the absorption, reflection and transmission of the ray are calculated based on the material properties and temperature conditions at the intersection; considering the heat transfer contribution of all rays comprehensively, the radiation heat transfer coefficient between the current unit and other units is calculated to form a radiation heat transfer coefficient matrix.

[0067] According to the material properties of the target mold steel, the radiation diffusion coefficient is preset to characterize the material's scattering and absorption capacity for radiation energy; for each finite element unit in the target mold steel, the unit center point or node is selected as the representative point to obtain the real-time temperature value of the point; in combination with the preset radiation diffusion coefficient and the representative point temperature, the mathematical model of radiation heat transfer, such as the P-1 model, is used to establish the relationship between radiation intensity and temperature; the radiation boundary conditions inside the furnace body, such as the radiation characteristics of the furnace wall and the radiation participation of the furnace gas, are considered to correct and supplement the calculation of radiation intensity; finally, the radiation intensity value of each finite element unit is determined as the input condition for the subsequent temperature field calculation.

[0068] In an optional embodiment, determining the radiation intensity according to a preset radiation diffusion coefficient and the temperature of any point in the finite element unit in the target mold steel in combination with a radiation model and a radiation boundary condition includes:

[0069] ;

[0070] in, represents the diffusion coefficient, κ represents the absorption coefficient, σ represents the Stefan-Boltzmann constant, T Indicates the material temperature, G Indicates the radiation intensity;

[0071] Radiation boundary conditions include:

[0072] ;

[0073] ε represents the emissivity of the material surface, T w represents the material temperature at the boundary of the finite element unit in the target mold steel, G w represents the radiation intensity at the boundary of the finite element unit in the target mold steel, n Represents the number of finite element elements.

[0074] In an optional embodiment, the method further comprises calculating the radiation heat transfer coefficient matrix according to a ray tracing algorithm:

[0075] Randomly emitting light inside the furnace where the target mold steel is located, the starting point, direction and energy of each light ray are randomly generated according to the temperature and emissivity of any point in the finite element unit of the target mold steel;

[0076] For each ray, the incident angle and reflection angle of the ray are calculated using the intersection of the ray and the finite element unit, and the absorptivity and reflectivity of the ray are calculated based on the material properties of the finite element unit;

[0077] Track the propagation path of the light inside the furnace until the light is completely absorbed or escapes from the furnace;

[0078] The light energy absorbed by each finite element unit is counted, and the radiation heat transfer coefficient matrix between finite element units is calculated.

[0079] In an optional implementation, calculating the radiation heat transfer coefficient matrix includes:

[0080] ;

[0081] in, Indicates the number of mold steel parts inside the furnace. i Finite element units to j The radiation heat transfer coefficient of the finite element unit is N Indicates the number of rays, θ ik , θ jk Respectively represent k The light and i and jThe angle between the normal vectors of the finite element, r ijk Indicates k The ray is in i and j The distance between the finite element units, V k Indicates k The reflectivity of the light, δ k Indicates k The absorption rate of a ray of light.

[0082] Exemplarily, according to the geometric shape of the target mold steel and the finite element division result, the spatial range inside the furnace body and the position information of each finite element unit are determined; for each finite element unit, the unit center point or node is selected as the potential starting point of the light; according to the temperature value of the finite element unit center point or node, the black body radiation law is used to calculate the radiation intensity and emissivity at the point; based on the radiation intensity and emissivity, the Monte Carlo method is used to randomly generate the starting point, direction and initial energy of the light, and the number of light rays is determined according to the set sampling accuracy; the above steps are repeated until a sufficient number of light rays are emitted inside the furnace body.

[0083] For each emitted light ray, the geometric relationship between the light ray and the boundary surface of the finite element unit is used to calculate the intersection of the light ray and each finite element unit; if the light ray has an intersection with a certain finite element unit, the position coordinates, incident angle and reflection angle of the intersection are recorded; based on the material properties of the finite element unit at the intersection, such as refractive index, absorption coefficient, etc., the absorptivity and reflectivity of the light ray in the unit are calculated; based on the absorptivity and reflectivity, the energy and direction of the light ray are updated, and the state information of the light ray is stored in the ray tracing list.

[0084] For each ray, initialize the ray tracing list according to its starting point, direction and energy; retrieve the status information of the current ray from the ray tracing list, including position, direction and energy; calculate the next intersection of the ray and the finite element unit inside the furnace according to the current position and direction of the ray; if the next intersection exists, calculate the absorptivity and reflectivity at the intersection, and update the energy and direction of the ray; if the energy of the ray is lower than a preset threshold, or the ray escapes from the furnace boundary, terminate the tracing of the current ray; store the updated ray status information in the ray tracing list until the ray is completely absorbed or escapes from the furnace.

[0085] For each finite element unit inside the furnace body, an energy absorption accumulator is initialized to count the total energy absorbed by the unit; the tracing results of all rays are traversed, and for each intersection of each ray, the light energy absorbed by the unit is calculated according to the absorption rate of the finite element unit at the intersection, and the energy value is accumulated in the corresponding energy absorption accumulator; after the ray tracing process is completed, the energy absorption accumulator of each finite element unit is normalized to obtain the proportion of total radiation energy absorbed by the unit; based on the energy absorption proportion, the radiation heat transfer coefficient between the finite element units is calculated to form a radiation heat transfer coefficient matrix, in which each element of the matrix represents the radiation heat transfer coefficient between two finite element units.

[0086] Through the above steps, the random emission of light inside the furnace where the target mold steel is located, the calculation of the intersection of light and finite element units, the tracking of light propagation paths, and the calculation of the radiation heat transfer coefficient matrix are realized. This method comprehensively considers the complex radiation heat transfer process inside the furnace, can more accurately describe the spatial distribution characteristics of radiation heat transfer, and provide important parameter support for subsequent temperature field analysis.

[0087] During the vacuum annealing process, due to the complex radiation environment inside the furnace, the calculation of the radiation heat source term is a key issue. Traditional radiation models, such as the Discrete Ordinates Method (DOM) and the Discrete Transfer Equation (DTE), can accurately describe the radiation heat transfer process, but the calculation complexity is high and it is difficult to effectively couple with the fluid flow and heat transfer equations. This application uses the P-1 radiation model to calculate the radiation heat source term, which can balance the calculation accuracy and efficiency of the radiation intensity.

[0088] The P-1 radiation model is a simplified radiation model based on spherical harmonic expansion. By truncating high-order terms, the radiation transfer equation is simplified to a diffusion equation. Compared with the traditional radiation model, the P-1 model greatly reduces the computational complexity while ensuring a certain accuracy, making it possible to solve the coupling of radiation with fluid flow and heat transfer. The P-1 model introduces the concept of radiation intensity, whose physical meaning is the radiation energy flux density within a unit solid angle. By solving the radiation intensity, the radiation heat source term can be obtained.

[0089] S102. Iteratively coupling the radiation heat transfer coefficient matrix and the radiation intensity, and obtaining the three-dimensional transient temperature field distribution inside the target mold steel by interpolation calculation;

[0090] For example, according to the geometric shape of the target mold steel and the finite element division results, an initial temperature value is assigned to each finite element unit to form an initial temperature field distribution; according to the initial temperature field distribution, the initial radiation intensity value of each finite element unit is calculated using the radiation intensity calculation method to form an initial radiation intensity field distribution. The calculated radiation heat transfer coefficient matrix is ​​introduced into the temperature field analysis as an important parameter for radiation heat transfer calculation; according to the radiation heat transfer coefficient matrix, the radiation heat transfer coupling relationship between finite element units is established to describe the transfer and distribution of radiation energy between units.

[0091] Select appropriate time steps and iteration termination conditions, such as the maximum number of iterations or temperature convergence accuracy; within each time step, based on the current temperature field distribution and radiation heat transfer coefficient matrix, calculate the net radiation heat flux of each finite element unit, that is, the difference between the absorbed radiation energy and the emitted radiation energy; apply the net radiation heat flux to the finite element unit as a heat load, combine the effects of other heat conduction and convection heat transfer, use the heat balance equation to calculate the temperature value of each finite element unit at the next moment, and update the temperature field distribution; according to the updated temperature field distribution, recalculate the radiation intensity value of each finite element unit, and update the radiation intensity field distribution until the iteration termination condition is met to obtain the converged temperature field and radiation intensity field distribution.

[0092] In each time step, the instantaneous temperature value at any position inside the target mold steel is calculated according to the temperature value of the finite element unit using interpolation methods such as linear interpolation, spline interpolation or Kriging interpolation. Through interpolation calculation in both space and time dimensions, a continuous and smooth three-dimensional transient temperature field distribution inside the target mold steel is obtained.

[0093] The three-dimensional transient temperature field distribution data obtained by interpolation calculation is visualized to generate temperature cloud maps, isothermal surface maps or dynamic evolution maps, etc., to intuitively display the spatial characteristics and time change trends of temperature distribution; quantitative analysis is performed on the temperature field distribution, and characteristic parameters such as temperature values, temperature gradients, and heating rates in key areas are extracted to evaluate the uniformity and rationality of temperature distribution; based on the analysis results of the temperature field distribution, the furnace structure design is optimized, process parameters are adjusted, or control strategies are improved to achieve a more optimized and stable temperature control effect.

[0094] Through the above steps, the iterative coupling of the radiation heat transfer coefficient matrix and the radiation intensity is realized, and the three-dimensional transient temperature field distribution inside the target mold steel is obtained by interpolation calculation. This method fully considers the mutual influence and dynamic change characteristics between radiation heat transfer and temperature field, and can more accurately and comprehensively describe the temperature distribution under complex working conditions.

[0095] S103. Calculate the temperature deviation according to the three-dimensional transient temperature field distribution and the target temperature setting value; take the temperature deviation as input, perform fuzzy reasoning through a preset fuzzy control rule library, and obtain adaptively adjusted PID controller parameters; substitute the adaptively adjusted PID controller parameters into the PID control algorithm, calculate the control output of the composite controller, and apply the control output to the heater of the target mold steel vacuum annealing furnace to achieve control of the annealing temperature.

[0096] In an optional implementation, the temperature deviation is used as input, and fuzzy reasoning is performed through a preset fuzzy control rule base to obtain the adaptively adjusted PID controller parameters including:

[0097] The temperature deviation is fuzzified by a triangular function with a variable width to obtain a temperature fuzzy value;

[0098] Through the preset fuzzy control rule base, the rule matching degree between the temperature fuzzy value and each fuzzy rule in the fuzzy control rule base is determined by the Mamdani reasoning method, and the corresponding matching weight is assigned to the rule matching degree, and the fuzzy rules are weightedly aggregated to obtain the reasoning result;

[0099] An adaptive scaling factor is introduced, and the inference result is defuzzified by using the maximum membership method to obtain adaptively adjusted PID controller parameters.

[0100] Exemplarily, based on the target temperature setting value and the actual measured temperature value, the temperature deviation at the current moment, that is, the difference between the target temperature and the actual temperature, is calculated; based on the temperature deviation at the current moment and the previous moment, the temperature deviation change rate, that is, the change in temperature deviation per unit time, is calculated.

[0101] According to the value range of temperature deviation, a set of fuzzy linguistic variables are designed, such as "negative large", "negative medium", "negative small", "zero", "positive small", "positive medium", "positive large", etc., to describe the fuzziness degree of temperature deviation; a triangular membership function with variable width is defined for each fuzzy linguistic variable, and the shape and parameters of the function are adaptively adjusted according to the distribution characteristics of temperature deviation; the temperature deviation at the current moment is taken as input, and its membership to each fuzzy linguistic variable is calculated to obtain the fuzzy representation of temperature deviation, namely the temperature fuzzy value.

[0102] Based on expert knowledge and control experience, a set of fuzzy control rules are established to describe the relationship between temperature deviation, deviation change rate and controller parameter adjustment; each fuzzy control rule consists of a premise part (IF) and a conclusion part (THEN), the premise part contains the fuzzy description of temperature deviation and deviation change rate, and the conclusion part contains the adjustment strategy of controller parameters; the fuzzy control rule can be in the form of "IF temperature deviation is A AND deviation change rate is B THEN controller parameter is adjusted to C", where A, B, C are corresponding fuzzy linguistic variables; according to the control objectives and process requirements, the fuzzy control rule base is designed and optimized to ensure the completeness, consistency and rationality of the rules.

[0103] The temperature fuzzy value and the deviation change rate fuzzy value are used as inputs to match each rule in the fuzzy control rule base; for each fuzzy control rule, the Mamdani reasoning method is used to calculate the matching degree between the temperature fuzzy value and the deviation change rate fuzzy value and the rule premise part; the matching degree is calculated through the T-norm (such as minimum value) or T-associative word (such as algebraic product) to obtain the rule matching degree of each rule; a weight value is assigned to the matching degree of each fuzzy control rule, and the weight value is set according to the importance and reliability of the rule.

[0104] The reasoning results of all fuzzy control rules are weightedly aggregated. Common aggregation methods include maximum method and weighted average method, etc., to obtain the aggregated fuzzy controller parameter adjustment amount; an adaptive scaling factor is introduced to scale the aggregated fuzzy controller parameter adjustment amount, and the adaptive scaling factor is adaptively adjusted according to the size of the temperature deviation and the deviation change rate; the scaled fuzzy controller parameter adjustment amount is defuzzified using the maximum membership method to convert the fuzzy amount into an accurate controller parameter adjustment value; the controller parameter adjustment value obtained by defuzzification is superimposed or replaced with the current PID controller parameters to obtain the adaptively adjusted PID controller parameters.

[0105] The adaptively adjusted PID controller parameters are updated to the controller, including the proportional coefficient, integral time constant and differential time constant, etc.; based on the updated controller parameters, the controller output is calculated, that is, the heating power of the target mold steel vacuum annealing furnace is adjusted; the controller output is applied to the heating system, and the annealing temperature is accurately controlled and dynamically tracked through power adjustment; continuous and real-time adaptive temperature control is achieved until the annealing process is completed.

[0106] Through the above steps, the temperature deviation and deviation change rate are used as inputs, and the PID controller parameters are adaptively adjusted through fuzzy processing, fuzzy reasoning, defuzzification and other operations, thereby realizing intelligent control of annealing temperature. This method combines the intelligence of fuzzy control and the stability of PID control, can effectively deal with parameter disturbances and uncertainties in the annealing process, and improve the robustness and adaptability of the control system.

[0107] In an optional implementation, before the temperature deviation is fuzzified by a triangular function of variable width to obtain a temperature fuzzy value, the method further includes constructing a triangular function of variable width:

[0108] The historical data of temperature deviation acquired in advance are clustered by fuzzy mean clustering algorithm to obtain multiple cluster centers;

[0109] The left width and right width of each cluster center are calculated according to the following formulas:

[0110] ;

[0111] ;

[0112] in, Indicates the left width, Indicates the right width, v c Indicates c Cluster centers, x h Indicates h Historical data, Indicates h Historical data for c The membership degree of the cluster centers is b represents the fuzzification parameter;

[0113] According to each cluster center, left width and right width, a variable width triangle function is constructed according to the following formula:

[0114] ;

[0115] in, μ c (x h ) Represents the value of the triangular function.

[0116] In an optional implementation, before substituting the adaptively adjusted PID controller parameters into the PID control algorithm to calculate the control output of the composite controller, the method further includes optimizing the controller parameters of the composite controller:

[0117] The controller parameters of the composite controller are optimized by particle swarm optimization algorithm;

[0118] The adaptively adjusted PID controller parameters are used as the particle positions, the particle speeds are defined, and the positions and speeds of a group of particles are randomly initialized to form an initial particle group;

[0119] According to the performance indicators of the control system, a fitness function is selected to evaluate the quality of the controller parameters corresponding to each particle;

[0120] In each iteration, the particle adjusts its speed and position according to its own optimal position and the global optimal position through the speed and position update formula. Inertia weight, acceleration constant and random number are introduced into the speed and position update formula. In the iterative process, the individual optimal solution and the global optimal solution of the particle are continuously updated until the preset termination condition is reached, and the optimized composite controller parameters are obtained.

[0121] A multi-group collaboration mechanism is introduced to divide the initial particle group into multiple sub-groups, and the collaboration and competition between sub-groups are realized through information exchange and migration operations.

[0122] Exemplarily, the adaptively adjusted PID controller parameters are used as the particle positions, including proportional coefficients, integral time constants, and differential time constants; the particle speed is defined to indicate the adjustment direction and step size of the controller parameters during the optimization process; the positions and speeds of a group of particles are randomly initialized to form an initial particle swarm, and the size of the particle swarm is set according to the complexity of the optimization problem and the limitations of computing resources.

[0123] According to the performance requirements of the control system, such as stability, response speed, overshoot, etc., a suitable fitness function is selected to evaluate the quality of the controller parameters corresponding to each particle; the fitness function can be constructed based on the integral index of the control error (such as ISE, IAE, ITAE, etc.), time domain performance index (such as rise time, adjustment time, overshoot, etc.) or frequency domain performance index (such as bandwidth, phase margin, gain margin, etc.); for each particle, its position information is substituted into the simulation model or actual system of the control system, and after running for a certain period of time, the fitness function value is calculated as the fitness evaluation of the particle.

[0124] In each iteration, for each particle, the new speed of the particle is calculated through the speed update formula according to its own historical best position (individual optimal solution) and the global best position (global optimal solution); in the speed update formula, inertia weight, acceleration constant and random number are introduced, among which the inertia weight controls the continuity of the particle speed, the acceleration constant controls the tendency of the particle to move toward the individual optimal solution and the global optimal solution, and the random number introduces the ability of random search; according to the updated speed, the new position of the particle, that is, the new value of the controller parameter, is calculated through the position update formula; the boundary processing of the updated particle position is performed to ensure that the controller parameters are within a reasonable value range.

[0125] For each particle, substitute its new position into the fitness function and calculate the new fitness value; compare the new fitness value of the particle with its historical best fitness value. If the new fitness value is better, update the individual optimal solution of the particle; compare the new fitness values ​​of all particles and find the optimal fitness value in the current iteration. If this value is better than the current global optimal solution, update the global optimal solution; until the preset number of iterations or the convergence condition of the fitness value is reached, the optimized controller parameters, that is, the global optimal solution, are obtained.

[0126] The initial particle swarm is divided into multiple subgroups, each of which performs particle swarm optimization independently, which is equivalent to multiple parallel optimization processes. During the iterative optimization process, information is exchanged between subgroups regularly, and the global optimal solution of each subgroup is shared with other subgroups to improve the global search capability of the optimization. The migration operation is introduced to migrate some particles in the subgroup with poor optimization performance to the subgroup with better performance to accelerate the optimization convergence. Through the collaboration and competition between subgroups, the search efficiency and global optimization capability of particle swarm optimization are improved to avoid premature convergence and local optimality.

[0127] The global optimal solution obtained by multi-group collaborative particle swarm optimization is used as the optimal parameters of the composite controller, including PID controller parameters and other controller parameters; the optimal controller parameters are substituted into the PID control algorithm and other control algorithms to calculate the control output of the composite controller; the output of the composite controller is applied to the heating system of the vacuum annealing furnace to achieve precise control and optimal adjustment of the annealing temperature.

[0128] Through the above steps, the composite controller parameters are optimized by particle swarm optimization algorithm based on the adaptively adjusted PID controller parameters, and the multi-group collaboration mechanism is introduced to improve the optimization efficiency and optimization ability, and finally the optimal composite controller parameters are obtained. This method combines the advantages of adaptive control, intelligent optimization and multi-group collaboration, and can achieve high-precision control and dynamic optimization of annealing temperature under complex working conditions, improving the performance and reliability of the control system.

[0129] Figure 2FIG. 1 is a schematic diagram of the structure of an intelligent control system for vacuum annealing of mold steel according to an embodiment of the present invention. Figure 2 As shown, the system comprises:

[0130] The first unit is used to divide the target mold steel and the interior of the furnace body where the target mold steel is located into a plurality of finite element units, and calculate the radiation heat transfer coefficient matrix between the finite element units inside the furnace body by using a ray tracing algorithm; and determine the radiation intensity according to a preset radiation diffusion coefficient and the temperature of any point in the finite element unit in the target mold steel, in combination with a radiation model and a radiation boundary condition;

[0131] The second unit is used for iteratively coupling the radiation heat transfer coefficient matrix and the radiation intensity, and obtaining the three-dimensional transient temperature field distribution inside the target mold steel through interpolation calculation;

[0132] The third unit is used to calculate the temperature deviation according to the three-dimensional transient temperature field distribution and the target temperature setting value; take the temperature deviation as input, perform fuzzy reasoning through a preset fuzzy control rule library, and obtain adaptively adjusted PID controller parameters; substitute the adaptively adjusted PID controller parameters into the PID control algorithm, calculate the control output of the composite controller, and apply the control output to the heater of the target mold steel vacuum annealing furnace to achieve control of the annealing temperature.

[0133] According to a third aspect of the embodiments of the present invention,

[0134] An electronic device is provided, comprising:

[0135] processor;

[0136] a memory for storing processor-executable instructions;

[0137] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0138] According to a fourth aspect of the embodiments of the present invention,

[0139] A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the aforementioned method is implemented.

[0140] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0141] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements 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.

Claims

1. Intelligent control method for vacuum annealing of mold steel, characterized in that: include: Divide the target mold steel and the interior of the furnace body where the target mold steel is located into a plurality of finite element units, and calculate the radiation heat transfer coefficient matrix between the finite element units inside the furnace body by using a ray tracing algorithm; Determining the radiation intensity according to a preset radiation diffusion coefficient and the temperature of any point in the finite element unit in the target mold steel in combination with a radiation model and a radiation boundary condition; Iteratively coupling the radiation heat transfer coefficient matrix and the radiation intensity, and obtaining the three-dimensional transient temperature field distribution inside the target mold steel through interpolation calculation; According to the three-dimensional transient temperature field distribution and the target temperature setting value, the temperature deviation is calculated; with the temperature deviation as input, fuzzy reasoning is performed through a preset fuzzy control rule library to obtain adaptively adjusted PID controller parameters; the adaptively adjusted PID controller parameters are substituted into the PID control algorithm to calculate the control output of the composite controller, and the control output is applied to the heater of the target mold steel vacuum annealing furnace to achieve control of the annealing temperature; According to the preset radiation diffusion coefficient and the temperature of any point in the finite element unit in the target mold steel, the radiation intensity is determined in combination with the radiation model and the radiation boundary condition, including: ； in, represents the diffusion coefficient, κ represents the absorption coefficient, σ represents the Stefan-Boltzmann constant, T Indicates the material temperature, G Indicates the radiation intensity; Radiation boundary conditions include: ； ε represents the emissivity of the material surface, T w represents the material temperature at the boundary of the finite element unit in the target mold steel, G w represents the radiation intensity at the boundary of the finite element unit in the target mold steel, n represents the number of finite element elements; The method further comprises calculating the radiation heat transfer coefficient matrix according to a ray tracing algorithm: Randomly emitting light inside the furnace where the target mold steel is located, the starting point, direction and energy of each light ray are randomly generated according to the temperature and emissivity of any point in the finite element unit in the target mold steel; For each ray, the incident angle and reflection angle of the ray are calculated using the intersection of the ray and the finite element unit, and the absorptivity and reflectivity of the ray are calculated based on the material properties of the finite element unit; Track the propagation path of the light inside the furnace until the light is completely absorbed or escapes from the furnace; Count the light energy absorbed by each finite element unit and calculate the radiation heat transfer coefficient matrix between finite element units; The calculation of the radiation heat transfer coefficient matrix includes: ； in, Indicates the number of mold steel parts inside the furnace. i Finite element units to j The radiation heat transfer coefficient of the finite element unit is N Indicates the number of rays, θ ik , θ jk Respectively represent k The light and i and j The angle between the normal vectors of the finite element, r ijk Indicates k The ray is in i and j The distance between the finite element units, V k Indicates k The reflectivity of the light, δ k Indicates k The absorption rate of the light beam; Taking the temperature deviation as input, fuzzy reasoning is performed through a preset fuzzy control rule base, and the adaptively adjusted PID controller parameters include: The temperature deviation is fuzzified by a triangular function with a variable width to obtain a temperature fuzzy value; Through the preset fuzzy control rule base, the rule matching degree between the temperature fuzzy value and each fuzzy rule in the fuzzy control rule base is determined by the Mamdani reasoning method, and the corresponding matching weight is assigned to the rule matching degree, and the fuzzy rules are weightedly aggregated to obtain the reasoning result; An adaptive scaling factor is introduced, and the inference result is defuzzified by using the maximum membership method to obtain adaptively adjusted PID controller parameters.

2. The method according to claim 1, characterized in that Before the temperature deviation is fuzzified by a triangular function of variable width to obtain a temperature fuzzy value, the method further includes constructing a triangular function of variable width: The historical data of temperature deviation acquired in advance are clustered by fuzzy mean clustering algorithm to obtain multiple cluster centers; The left width and right width of each cluster center are calculated according to the following formulas: ； ； in, Indicates the left width, Indicates the right width, v c Indicates c Cluster centers, x h Indicates h Historical data, Indicates h Historical data for c The membership degree of the cluster centers is b represents the fuzzification parameter; According to each cluster center, left width and right width, a variable width triangle function is constructed according to the following formula: ； in, μ c (x h ) Represents the value of the triangular function.

3. The method according to claim 1, characterized in that Substituting the adaptively adjusted PID controller parameters into the PID control algorithm to calculate the control output of the composite controller, the method further includes optimizing the controller parameters of the composite controller: The controller parameters of the composite controller are optimized by particle swarm optimization algorithm; The adaptively adjusted PID controller parameters are used as the particle positions, the particle speeds are defined, and the positions and speeds of a group of particles are randomly initialized to form an initial particle group; According to the performance indicators of the control system, a fitness function is selected to evaluate the quality of the controller parameters corresponding to each particle; In each iteration, the particle adjusts its speed and position according to its own optimal position and the global optimal position through the speed and position update formula. Inertia weight, acceleration constant and random number are introduced into the speed and position update formula. In the iterative process, the individual optimal solution and the global optimal solution of the particle are continuously updated until the preset termination condition is reached, and the optimized composite controller parameters are obtained. A multi-group collaboration mechanism is introduced to divide the initial particle group into multiple sub-groups, and the collaboration and competition between sub-groups are realized through information exchange and migration operations.

4. An intelligent control system for vacuum annealing of mold steel, used to implement the method described in any one of claims 1 to 3, characterized in that: include: The first unit is used to divide the target mold steel and the interior of the furnace body where the target mold steel is located into a plurality of finite element units, and calculate the radiation heat transfer coefficient matrix between the finite element units inside the furnace body by using a ray tracing algorithm; Determining the radiation intensity according to a preset radiation diffusion coefficient and the temperature of any point in the finite element unit in the target mold steel in combination with a radiation model and a radiation boundary condition; The second unit is used for iteratively coupling the radiation heat transfer coefficient matrix and the radiation intensity, and obtaining the three-dimensional transient temperature field distribution inside the target mold steel through interpolation calculation; The third unit is used to calculate the temperature deviation according to the three-dimensional transient temperature field distribution inside the target mold steel and the target temperature setting value; take the temperature deviation as input, perform fuzzy reasoning through a preset fuzzy control rule library, and obtain the adaptively adjusted PID controller parameters; substitute the adaptively adjusted PID controller parameters into the PID control algorithm, calculate the control output of the composite controller, and apply the control output to the heater of the vacuum annealing furnace of the target mold steel to achieve control of the annealing temperature.

5. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method described in any one of claims 1 to 3.

6. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 3 is implemented.

Images