Decision-making control system and method based on temperature field and atmosphere analysis

Abstract

The present invention discloses a decision-making control system and method based on temperature field and atmosphere analysis, which relates to the field of intelligent control technology. In the system, a monitoring module obtains a temperature set and a gas concentration set; a trend analysis module constructs a three-dimensional model and establishes a coordinate system, a variation function combines with a Kriging equation to supplement the temperature set to generate an extended temperature set, a discrete three-dimensional model is a plurality of nodes and is matched with points in the extended temperature set through nearest neighbor search, a heat conduction equation and a convection boundary are set and a finite element matrix equation group is constructed, the temperature field model is obtained by solving and a temperature trend curve is statistically generated; a decision module compares the temperature trend curve, the gas concentration set and the process standard to obtain an abnormal segment number, and adopts a fuzzy PID control algorithm to analyze the segment temperature difference and the concentration error vector of the abnormal segment number to generate an adjustment vector; an execution module automatically adjusts based on the adjustment vector, and realizes intelligent decision adjustment based on the joint analysis of the temperature field and the atmosphere.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, and in particular to a decision-making control system and method based on temperature field and atmosphere analysis. Background Art

[0002] In modern industrial production, many processes rely heavily on precise control of the temperature field and atmosphere of functional equipment. Precise temperature and a suitable gas environment not only affect product quality but also directly impact production efficiency and equipment lifespan. Therefore, the development of efficient monitoring and control systems is imperative.

[0003] Traditional monitoring and control systems have flaws. They often perform trend analysis and intelligent control based on temperature data or gas composition collected by sensors. These systems lack the ability to analyze spatial temperature variations, resulting in inaccurate temperature trend analysis and ignoring the impact of temperature fields and atmosphere on industrial production. This leads to one-sided control methods. Summary of the Invention

[0004] In response to the deficiencies in the prior art, the present invention proposes a decision-making control system and method based on temperature field and atmosphere analysis, which realizes the fine modeling of the temperature field model and intelligent decision-making adjustment based on temperature field and atmosphere analysis.

[0005] The technical solution to achieve the purpose of the present invention is:

[0006] Decision-making control system based on temperature field and atmosphere analysis, including monitoring module, trend analysis module, decision module and execution module;

[0007] The monitoring module sets the monitoring time periodically, and obtains the temperature set through the temperature measurement equipment group and gas detection equipment at each monitoring time. and gas concentration sets , where the temperature set and gas concentration sets Included The temperature sequence of each functional segment and The gas concentration vector of each functional segment, is the total number of functional segments;

[0008] The trend analysis module builds a three-dimensional model of the functional equipment and establishes a Cartesian space coordinate system. It uses the variation function to describe the spatial variation characteristics of the temperature and uses the Kriging equation to analyze the temperature set. The measured temperature in the space is expanded to generate the expanded temperature set , the mesh segmentation technology is used to simplify and discretize the three-dimensional model into nodes, expand the temperature set through nearest neighbor search The temperature of each point in the graph is assigned to the surface node that matches the point coordinates, and the ambient temperature is obtained. , based on the thermal conductivity of functional devices Set up the heat conduction equation and define the convection boundary , the finite element method is used to construct the finite element matrix equation and the conjugate gradient method is used to solve the temperature field model , for the temperature field model exist Statistical analysis is performed in the axial direction to generate temperature trend curves ,in, is the total number of nodes, temperature field model include The temperature of each node, temperature trend curve reflect The temperature change trend of each functional segment, for Coordinate variables in the axis direction;

[0009] The decision module converts the temperature trend curve and gas concentration sets Compare with the corresponding process standards to obtain the abnormal segment number, take the segment temperature difference and concentration error vector of each abnormal segment number as the input of the fuzzy PID control algorithm, use fuzzy logic to dynamically adjust the PID control algorithm parameters, and output the adjustment vector corresponding to each abnormal segment number. The process standard includes the standard temperature trend curve and gas standard concentration sets , the adjustment vector includes the fuel valve adjustment amount and the air valve adjustment amount;

[0010] The execution module automatically adjusts the heating equipment of the corresponding functional segment based on the adjustment vector of each abnormal segment number.

[0011] Furthermore, the monitoring module includes a temperature measurement unit and a concentration detection unit;

[0012] Temperature measurement unit acquisition The measured temperatures of the temperature measuring devices with the same segment number are sorted according to the device number sequence to build a temperature sequence, and the temperature sequence is integrated. The temperature sequence of each functional segment is used to construct a temperature set And transmit to the trend analysis module, is the total number of functional segments, is the total number of temperature measuring devices;

[0013] Concentration detection unit acquisition The gas concentration vector measured by the gas detection equipment in each functional segment is integrated The gas concentration vectors of the functional segments are used to construct the gas concentration set And transmitted to the decision module, where the Gas concentration vector of each functional segment Including The oxygen concentration, carbon dioxide concentration, carbon monoxide concentration and nitrogen concentration of each functional section.

[0014] Specifically, the gas detection equipment uses a semiconductor laser gas analyzer, which uses multiple semiconductor lasers of different wavelengths to emit lasers that match the specific absorption lines of oxygen, carbon dioxide, carbon monoxide and nitrogen. By detecting the laser intensity of each wavelength after passing through the mixed gas, the oxygen concentration, carbon dioxide concentration, carbon monoxide concentration and nitrogen concentration are determined according to the Beer-Lambert law.

[0015] Specifically, the three-dimensional model of the functional equipment needs to be based on The geometric parameters of each functional segment are pre-constructed through 3D modeling methods, which include surface modeling, point cloud reverse modeling, image reverse modeling and 3D scanning modeling. The central axis of the 3D model is used as the Axes to construct a Cartesian space coordinate system.

[0016] Furthermore, the variation function is used to describe the spatial variation characteristics of temperature, and the temperature set is The measured temperature in the space is expanded to generate the expanded temperature set , including the following specific steps:

[0017] Defining Temperature Sets middle The temperature measuring device coordinates of the temperature measuring device are , will Temperature measuring device coordinates Hedi Temperature measuring device coordinates Temperature measuring equipment And calculate the corresponding spatial distance , a total of A pair of temperature measuring devices and the corresponding spatial distance, , For the The temperature measuring device coordinates of each temperature measuring device;

[0018] Calculate the temperature measurement equipment The variance function value of , the variation function value For the Temperature measuring device coordinates Hedi Temperature measuring device coordinates The semivariance of the temperature at , used to quantitatively describe the spatial distance The variability of the spatial distance is obtained by fitting the existing spatial distance and the variability function value through the least squares method. ;

[0019] Based on the unbiased and minimum variance estimation principles, the coordinates to be estimated are Establish the Kriging equation system to obtain the optimal weight coefficient vector , the optimal weight coefficient vector Satisfy the coordinates of the point to be estimated With the Temperature measuring device coordinates The variance function value is equal to the optimal weight coefficient vector With the Temperature measuring device coordinates The sequence of variance function values The inner product of the optimal weight coefficient vector The optimal weight coefficient sum is 1;

[0020] Solve to obtain the coordinates of the point to be estimated The optimal weight coefficient vector , coordinates of the point to be estimated Estimated temperature Temperature set in The known temperature vector and the optimal weight coefficient vector are arranged into measured temperatures The inner product of

[0021] Uniform selection on the surface of a 3D model points to be estimated and calculate the estimated temperature , and the temperature set Merge to generate extended temperature set , expand the temperature set Including functional device surfaces The temperature of a point, is the total number of points.

[0022] Specifically, the grid segmentation technology removes the detailed structures that have little influence on the temperature field modeling in the three-dimensional model, generates a simplified three-dimensional model, and translates each coordinate axis of the coordinate system along the direction of the other two coordinate axes to generate a parallel straight line group. The straight lines in the parallel straight line group of the three coordinate axes intersect each other, and the intersection of any two straight lines is defined as a node, and a total of nodes and define the empty hexahedron surrounded by 8 nodes as a unit, and obtain a total of units, is the total number of units, for the expanded temperature set The temperature of each point in the surface is obtained by performing a nearest neighbor search based on the point coordinates, retrieving the node with the smallest spatial distance from the point coordinates and assigning the temperature of the corresponding point. All nodes that have been assigned temperatures are defined as known surface nodes.

[0023] Furthermore, based on the thermal conductivity of the functional device Set up the heat conduction equation and define the convection boundary , the finite element method is used to construct the finite element matrix equation and the conjugate gradient method is used to solve the temperature field model , including the following specific steps:

[0024] Determine thermal conductivity , set up the heat conduction equation to describe the heat flux density With temperature gradient The ratio is equal to the thermal conductivity The opposite of

[0025] Defining Convection Boundaries , the heat flux density of convective heat transfer is described by Newton's law of cooling and surface temperature and ambient temperature The difference is proportional to the convective heat transfer coefficient. ;

[0026] consider The thermal balance inside each unit is calculated based on Fourier's law. The heat conduction equation in each unit is analyzed and the weighted residual method is used to construct the The finite element equations of the unit, the finite element equations describe the The second-order temperature gradient within the unit volume Integrals on the cell boundary Heat flux At the unit boundary Area The integral on the graph shows a proportional relationship, which is equal to the thermal conductivity , ;

[0027] Use the The interpolation of the shape function of each node in the element describes the The spatial coordinates within a unit are Temperature indication , the shape function is determined according to the geometric characteristics and node distribution of the element, 、 and They are Axis direction, Axis direction and Coordinate variables in the axis direction;

[0028] The first The temperature of each unit is represented Substitute The finite element equation of each unit is derived through integration operation and matrix transformation and organized into a finite element matrix equation to describe the The stiffness matrix of each element and the node temperature vector The product of is equal to the load vector ;

[0029] Will The finite element matrix equations of each unit are updated according to the assembly rules. The assembly rules are established based on the node sharing relationship of adjacent units. The values ​​of the common nodes in the stiffness matrices and load vectors of adjacent units are accumulated and replaced accordingly, integrated into a finite element matrix equation group and solved using the conjugate gradient method to obtain The node temperature vector of each element generates a temperature field model .

[0030] Specifically, for the temperature field model exist Statistical analysis in the axial direction can reflect the temperature change trend of different functional segments, and the temperature field model Same coordinate variables The temperature is averaged to generate coordinate variables The corresponding average temperature, traversing all coordinate variables To obtain the corresponding average temperature, the temperature trend curve is obtained using least squares fitting. .

[0031] Furthermore, compare the temperature trend curve Trend curve with standard temperature Obtaining the exception segment number includes the following specific steps:

[0032] Determined based on the 3D model Functional segments in Segment coordinate variable range in the axis direction;

[0033] Calculate temperature trend curve Trend curve with standard temperature In the Temperature mean square error of each functional segment , No. Temperature mean square error of each functional segment Temperature trend curve Trend curve with standard temperature The square of the error in Segment coordinate variable range in the axis direction The definite integral within is divided by the length of the segment coordinate variable , and Respectively Functional segments in The starting and ending coordinate variables of the axis direction, ;

[0034] If the temperature mean square error is greater than or equal to the temperature mean square error threshold, Marked as abnormal segment number, if the temperature mean square error If the error is less than the temperature mean square error threshold, no processing is performed.

[0035] Furthermore, the gas concentration Gas standard concentration set To obtain the exception segment number, the following specific steps are included:

[0036] Get gas concentration set Gas standard concentration set Middle Gas concentration vector of each functional segment and the gas standard concentration vector ;

[0037] Calculate the The mean square error of the concentration of each functional segment , concentration mean square error is the concentration error vector The square of the second norm , No. The concentration error vector of the functional segment Equal to the gas concentration vector Subtract the gas standard concentration vector ;

[0038] If the concentration mean square error is greater than or equal to the concentration mean square error threshold, Marked as abnormal segment number, if the concentration mean square error If the value is less than the concentration mean square error threshold, no processing is performed.

[0039] Specifically, the temperature difference of the abnormal functional section is the temperature trend curve Trend curve with standard temperature The definite integral of the error within the corresponding segment independent variable range is divided by the segment independent variable length. The concentration error vector of the abnormal functional segment is the gas concentration vector of the abnormal functional segment minus the gas standard concentration vector. The concentration error vector includes oxygen concentration error, carbon dioxide concentration error, carbon monoxide concentration error and nitrogen concentration error.

[0040] Specifically, the fuzzy PID control algorithm includes fuzzy reasoning. The input variables of the fuzzy reasoning include the temperature difference of the abnormal functional section, the oxygen concentration error, the carbon dioxide concentration error, the carbon monoxide concentration error and the nitrogen concentration error. The output variables of the fuzzy reasoning include the proportional factor, integral factor and differential factor of the PID control algorithm. The domain of each input variable or output variable is determined, and the fuzzy sets are defined to include "negative large", "negative small", "zero", "positive medium" and "positive large". Fuzzy rules are formulated based on experience. The fuzzy rules establish the correspondence between the input fuzzy set group and the output fuzzy set group. For each fuzzy rule, the membership of each input variable to the corresponding fuzzy set in the input fuzzy set group is calculated by the Gaussian membership function. The activation strength of the fuzzy rule is obtained by AND operation. The activation strengths of all fuzzy rules are ANDed to determine the membership of the output fuzzy set group and each output variable. The center of gravity method is used for defuzzification, and the proportional factor, integral factor and differential factor of the PID control algorithm are output.

[0041] Specifically, the fuzzy PID control algorithm includes a PID control algorithm, which performs weighted summation of the temperature difference, oxygen concentration error, carbon dioxide concentration error, carbon monoxide concentration error and nitrogen concentration error of the abnormal functional segment to generate a comprehensive error, obtains a proportional factor, an integral factor and a differential factor, multiplies the comprehensive error by the proportional factor to generate a proportional control amount and determines the residual comprehensive error, multiplies the integral operation result of the residual comprehensive error to generate an integral control amount and updates the residual comprehensive error, repeats the integral adjustment until the residual comprehensive error is less than the comprehensive error threshold and then stops, and the differential adjustment multiplies the comprehensive error change rate of two adjacent integral adjustments by the differential factor to generate a differential correction control amount, the number of differential adjustments is the same as the number of integral adjustments, and the sum of the proportional control amount, multiple integral control amounts and multiple differential control amounts is used as the control amount, and the fuel valve adjustment amount and the air valve adjustment amount are obtained by considering the stoichiometric ratio and combined into an adjustment vector corresponding to the abnormal functional segment.

[0042] The decision-making control method based on temperature field and atmosphere analysis and the decision-making control system based on temperature field and atmosphere analysis are implemented, including the following specific steps:

[0043] At each monitoring moment, the temperature set is obtained by monitoring the temperature measurement equipment group and gas detection equipment. and gas concentration sets ;

[0044] based on The geometric parameters of each functional segment are used to construct a three-dimensional model and establish a Cartesian space coordinate system. is the total number of functional segments;

[0045] The variation function is used to describe the spatial variation characteristics of temperature, and the temperature set is The measured temperature in the space is expanded to generate the expanded temperature set ;

[0046] The mesh segmentation technique is used to simplify and discretize the three-dimensional model into nodes, expand the temperature set through nearest neighbor search The temperature of each point in the graph is assigned to the corresponding node with matching coordinates. is the total number of nodes;

[0047] Get the ambient temperature , based on the thermal conductivity of functional devices Set up the heat conduction equation and define the convection boundary , construct the finite element matrix equations by the finite element method and solve them by the conjugate gradient method to obtain the temperature field model ;

[0048] Temperature field model exist Statistical analysis is performed in the axial direction to generate temperature trend curves And with the standard temperature trend curve Compare and combine the gas concentration Gas standard concentration set The comparison result determines the abnormal segment number;

[0049] The temperature difference and concentration error vector of each abnormal segment number are used as the input of the fuzzy PID control algorithm. The fuzzy logic is used to dynamically adjust the PID control algorithm parameters, and the adjustment vector corresponding to each abnormal segment number is output and adjusted.

[0050] Compared with the prior art, the present invention constructs a three-dimensional model of the functional equipment and establishes a coordinate system, uses a variation function to describe the spatial variation of the temperature in the temperature set on the surface of the functional equipment, supplements and generates an expanded temperature set through the Kriging equation to support temperature field modeling, simplifies the three-dimensional model into several nodes and discretizes them into several nodes and matches them with the points in the expanded temperature set through nearest neighbor search, sets the heat conduction equation based on thermal conductivity and defines the convection boundary in combination with the ambient temperature, uses the finite element method to construct a finite element matrix equation group, solves the temperature of the unmatched nodes through the conjugate gradient method, constructs a detailed temperature field model of the functional equipment and fits to obtain the temperature trend curve, providing accurate data support for subsequent intelligent decision-making; compares the temperature trend curve and the gas concentration set with the process standard to determine the abnormal functional segment, and uses the fuzzy PID control algorithm to analyze the segment temperature difference and concentration error vector of the abnormal functional segment to generate the corresponding adjustment vector and make real-time adjustments, thereby realizing accurate analysis of the temperature trend and intelligent decision-making adjustment based on the joint analysis of the temperature field and atmosphere. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1This is a schematic diagram of the principle structure of the decision-making control system based on temperature field and atmosphere analysis in the present invention;

[0052] Figure 2 A flow chart for generating an expanded temperature set in the present invention;

[0053] Figure 3 A flow chart for obtaining abnormal segment numbers by comparing a gas concentration set with a gas standard concentration set in the present invention;

[0054] Figure 4 This is a flow chart of the decision-making control method based on temperature field and atmosphere analysis in the present invention. DETAILED DESCRIPTION

[0055] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0056] Example 1

[0057] like Figure 1 As shown, a specific embodiment of the present invention discloses a decision control system based on temperature field and atmosphere analysis, including a monitoring module, a trend analysis module, a decision module and an execution module;

[0058] Monitoring module to monitor intervals The monitoring time is set periodically, and the temperature set is obtained at each monitoring time through the segmented temperature measurement equipment group and gas detection equipment monitoring. and gas concentration sets , where the temperature set and gas concentration sets Included The temperature sequence of each functional segment and The gas concentration vector of each functional segment, is the total number of functional sections. For example, the functional sections of a continuous annealing furnace include a heating section, a holding section, and a slow cooling section. The corresponding section numbers can be marked as 1, 2, and 3 respectively.

[0059] Trend analysis module is based on The geometric parameters of each functional segment are used to construct a three-dimensional model of the functional equipment and establish a Cartesian space coordinate system. The variation function is used to describe the spatial variation characteristics of the temperature in the functional equipment. The temperature set is calculated based on the Kriging equation. The measured temperature in the 3D model is spatially expanded on the surface to generate an expanded temperature set. , the mesh segmentation technology is used to simplify and discretize the three-dimensional model into nodes, expand the temperature set through nearest neighbor search The temperature of each point in the graph is assigned to the corresponding surface node that matches the point coordinates, and the ambient temperature is obtained. , based on the thermal conductivity of functional devices Set up the heat conduction equation and define the convection boundary , construct the finite element matrix equation through the finite element method, and solve the finite element matrix equation through the conjugate gradient method to obtain the temperature field model , for the temperature field model exist Statistical analysis is performed in the axial direction to generate temperature trend curves ,in, is the total number of nodes. The coordinates of the temperature measuring equipment are determined according to the installation parameters of the temperature measuring equipment and the Cartesian space coordinate system. The temperature field model include The temperature of each node, temperature trend curve reflect The temperature change trend of each functional segment, for Coordinate variables in the axial direction, for example, the temperature trend curve of a continuous annealing furnace Reflects the temperature change trend of the heating section, insulation section and slow cooling section;

[0060] The decision module compares the temperature trend curve Trend curve with standard temperature and comparison gas concentration sets Gas standard concentration set To obtain the abnormal segment number, the segment temperature difference and concentration error vector of each abnormal segment number is used as the input of the fuzzy PID control algorithm, and the PID control algorithm parameters are dynamically adjusted using fuzzy logic to output the adjustment vector corresponding to each abnormal segment number. Among them, the standard temperature trend curve and gas standard concentration sets The process standards predefined for functional equipment include fuel valve adjustment and air valve adjustment. For example, the heating, holding, and slow cooling sections of a continuous annealing furnace are all heated by burning natural gas. In this case, the fuel valve adjustment and air valve adjustment are actually adjustments to the valve opening, used to control the combustion ratio of natural gas and air.

[0061] The execution module automatically adjusts the fuel valve opening and air valve opening of the heating equipment of the corresponding functional segment based on the adjustment vector of each abnormal segment number. For example, the heating equipment of the heating section, insulation section and slow cooling section of the continuous annealing furnace includes a double regenerative burner and a high-speed burner burner in the heating section.

[0062] Furthermore, the monitoring module includes a temperature measurement unit and a concentration detection unit;

[0063] Temperature measurement unit acquisition The measured temperature of each temperature measuring device in each functional segment is sorted into a temperature sequence according to the device number sequence with the same segment number. The temperature sequence of each functional segment is used to construct a temperature set And transmitted to the trend analysis module, where For the The temperature sequence of each functional segment, For the The temperature measurement group number in the temperature measurement equipment group of each functional segment is The measured temperature collected by the temperature measuring equipment, is the corresponding device number, and Respectively before Functional segments and front The maximum device number counted in each function segment, , , For the The maximum temperature measurement group number of the temperature measurement equipment group in each functional segment;

[0064] Concentration detection unit acquisition The gas concentration vector measured by the gas detection equipment in each functional segment is integrated The gas concentration vectors of the functional segments are used to construct the gas concentration set And transmitted to the decision module, where For the The gas concentration vector of each functional segment, 、 、 and Respectively The gas detection equipment of each functional segment measures the oxygen concentration, carbon dioxide concentration, carbon monoxide concentration and nitrogen concentration. The equipment number of the gas detection equipment in each functional segment is , is the total number of temperature measuring devices, ,therefore, The temperature measuring equipment of the functional segment has occupied the equipment number 1 to , the equipment number of the gas detection equipment needs to be from Initially, each functional section requires a gas detection device, and the concentration detection unit includes Gas detection equipment, equipment number occupies to .

[0065] Specifically, when applied to a continuous annealing furnace, it is known that the section numbers corresponding to the heating section, the holding section, and the slow cooling section are 1, 2, and 3 respectively, and 4 temperature sensors are set at the sockets of the heating section, the holding section, and the slow cooling section respectively. Then the total number of temperature measurement group numbers corresponding to the 1st to 3rd functional sections is 、 and All are 4. In this case, the temperature measurement unit includes 12 temperature measurement devices, namely , the corresponding equipment numbers occupy 1 to 12. Since the heating section, insulation section and slow cooling section all require gas detection equipment, the concentration detection unit includes 3 gas detection equipment in total, and the corresponding equipment numbers occupy 13 to 15. Taking the temperature sensor with the temperature measurement group number 3 in the temperature measurement equipment group of the third functional section as an example, the corresponding equipment number is , and the gas detection equipment corresponding to the second functional segment is numbered .

[0066] Specifically, the gas detection equipment uses a semiconductor laser gas analyzer, which uses multiple semiconductor lasers of different wavelengths to emit lasers that precisely match the specific absorption spectra of oxygen, carbon dioxide, carbon monoxide and nitrogen. When the laser passes through a mixed gas including these gases, each gas absorbs the laser of the corresponding wavelength, causing the laser intensity to weaken. By detecting the changes in the laser intensity of each wavelength, the oxygen concentration, carbon dioxide concentration, carbon monoxide concentration and nitrogen concentration are determined respectively according to the Beer-Lambert law.

[0067] Specifically, the three-dimensional model of the functional equipment needs to be based on The geometric parameters of each functional segment are pre-constructed through the 3D modeling method. The 3D model can accurately reflect the actual physical structure of the functional equipment. The 3D modeling method includes surface modeling, point cloud reverse modeling, image reverse modeling and 3D scanning modeling. After obtaining the 3D model, a Cartesian space coordinate system is established. The axis is limited to the central axis of the 3D model, and the other two coordinate axes and the origin can be defined by yourself.

[0068] like Figure 2 As shown in the figure, further, the variation function is used to describe the spatial variation characteristics of temperature in functional equipment, and the temperature set is The measured temperature in the 3D model is spatially expanded on the surface to generate an expanded temperature set. , including the following specific steps:

[0069] Defining Temperature Sets middle The temperature measuring device coordinates of the temperature measuring device are , will Temperature measuring device coordinates Hedi Temperature measuring device coordinates Temperature measuring equipment And calculate the corresponding spatial distance , a total of A pair of temperature measuring devices and the corresponding spatial distance, ,The coordinates of the temperature measuring device are determined based on the installation parameters of the temperature measuring device and the Cartesian space coordinate system;

[0070] Calculate the temperature measurement equipment The variance function value of , the variation function value For the Temperature measuring device coordinates Hedi Temperature measuring device coordinates The semivariance of the temperature at , used to quantitatively describe the spatial distance Variability, variogram value The larger the distance, the The more drastic the temperature change under the scale, the greater the value of the variation function. The specific formula is as follows:

[0071] ;

[0072] in, and Temperature measuring equipment Middle Temperature measuring device coordinates Hedi Temperature measuring device coordinates The temperature at the location is obtained by fitting a polynomial using the least squares method based on the existing spatial distance and variogram value. ;

[0073] The coordinates to be estimated Establish the Kriging equation system and solve the optimal weight coefficient vector based on the unbiased and minimum variance estimation principles , the optimal weight coefficient vector Satisfy the coordinates of the point to be estimated With the Temperature measuring device coordinates The variance function value is equal to the optimal weight coefficient vector With the Temperature measuring device coordinates The sequence of variance function values The inner product of Including Temperature measuring device coordinates With temperature set middle The variation function value of the temperature measuring device coordinates, that is, , and the optimal weight coefficient vector The sum of the sequence is 1, so the coordinates of the point to be estimated The Kriging equations include The equations are organized into matrix form as follows:

[0074] ;

[0075] in, The coordinates of the points to be estimated are and The variation function value corresponding to the spatial distance of the temperature measurement device coordinates, They are the first optimal weight to the Optimal weights;

[0076] Solve to obtain the coordinates of the point to be estimated The optimal weight coefficient vector , then the coordinates of the point to be estimated are Estimated temperature Temperature set middle The known temperature vector and the optimal weight coefficient vector are arranged into the measured temperatures of the temperature measuring devices The inner product of ;

[0077] Since the temperature measuring equipment is installed on the surface of the functional equipment, the surface of the three-dimensional model of the functional equipment is uniformly selected. points to be estimated and estimate the corresponding estimated temperature , and the temperature set Merge to generate extended temperature set , expand the temperature set Including functional device surfaces The temperature of a point, is the total number of points.

[0078] Specifically, the mesh segmentation technology first simplifies the three-dimensional model of the functional equipment, removes the detailed structure in the three-dimensional model that has little influence on the temperature field modeling, generates a simplified three-dimensional model, and translates each coordinate axis of the Cartesian space coordinate system along the other two coordinate axes to generate a group of parallel straight lines. The straight lines in the parallel straight line group of the three coordinate axes intersect each other to cut the simplified three-dimensional model. The intersection of any two straight lines is a node, and the discrete generation nodes, define the empty hexahedron surrounded by 8 nodes as a unit, then Nodes can be built units, where an empty hexahedron means a hexahedron with no nodes inside. and are the total number of nodes and the total number of elements, respectively, and the total number of nodes Much larger than the total number of points The distance between two adjacent parallel lines in a parallel line group is defined based on the spatial distance between each position and the heat source in the simplified three-dimensional model. For example, the distance between adjacent parallel lines close to the burner in the heating section, holding section and slow cooling section of a continuous annealing furnace is defined as the distance between adjacent parallel lines away from the burner. , for the expanded temperature set The temperature of each point in the surface is obtained by performing a nearest neighbor search based on the corresponding point coordinates, retrieving the node with the smallest spatial distance from the point coordinates and assigning the temperature of the corresponding point. All nodes that have been assigned temperatures are defined as known surface nodes.

[0079] Furthermore, based on the thermal conductivity of the functional device Set up the heat conduction equation and define the convection boundary , construct the finite element equation by finite element method and solve it by conjugate gradient method to obtain the temperature field model , including the following specific steps:

[0080] Determine thermal conductivity based on the material of the functional device , set up the heat conduction equation to describe the heat flux density With temperature gradient The ratio is equal to the thermal conductivity The opposite number of ;

[0081] Defining Convection Boundaries , considering the convection heat transfer between the surface of the functional device and the ambient air, the heat flux density of the convection heat transfer is described by Newton's cooling law and surface temperature and ambient temperature The relationship between the convection boundary On, satisfied , is the convective heat transfer coefficient, which reflects the convective heat transfer intensity between the surface and the ambient air. Includes temperatures of all known surface nodes;

[0082] consider Nodes built The heat balance inside each unit is The heat conduction equation within each unit is analyzed. Based on Fourier's law, heat is transferred from high temperature to low temperature, and the transfer rate is related to the temperature gradient. Proportional to the first The finite element equations of each unit are as follows:

[0083] ;

[0084] in, represents the volume variable, and Respectively The cell volume and cell boundary of each cell, and Unit boundaries The heat flux and area, 、 and Respectively The temperature within the unit In the Cartesian coordinate system Axis direction, Axis direction and The second-order temperature gradient in the axial direction reflects the temperature gradient changes in ;

[0085] Use the The interpolation of the shape function of each node in the element describes the The spatial coordinates within a unit are Temperature indication , 、 and They are Axis direction, Axis direction and The coordinate variables in the axis direction are as follows:

[0086] ;

[0087] in, and Respectively Unit No. The shape function and node temperature of each node, shape function , 、 and Respectively Unit No. The node coordinates of the nodes, ;

[0088] The first The temperature of each unit is represented Substitute The finite element equation of each unit is deduced through integration operation and matrix transformation and organized into a finite element matrix equation in matrix form. ,in, 、 and Respectively The stiffness matrix, node temperature vector and load vector of each element, the stiffness matrix and load vector Used to describe the The influence of the temperature of the eight nodes in a unit and the boundary heat flow on the The influence of the 8 node temperatures within a unit;

[0089] Will The finite element matrix equation of each unit is updated according to the assembly rule. The assembly rule is established based on the node sharing relationship of adjacent units. Unit and When the units are adjacent, the Unit and There are 4 common nodes in each unit. For a single common node, extract the The stiffness matrix of each element Hedi The stiffness matrix of each element The influence vectors related to a single common node in the CNN are accumulated and replaced to extract the first Load vector for each element With the Load vector for each element The influence associated with a single common node is replaced by cumulatively. The finite element matrix equations of each unit are integrated into the finite element matrix equation group after updating according to the assembly rules, and the conjugate gradient method is used to solve it. The node temperature vector of each unit is assigned to the corresponding node to generate a temperature field model. .

[0090] Specifically, due to the Cartesian coordinate system The axis is the central axis of the functional device, which runs through all functional sections of the functional device, so the temperature field model exist Statistical analysis in the axial direction can effectively reflect the temperature change trend of different functional segments. exist The axis directions have the same coordinate variables The temperature is averaged to generate coordinate variables The corresponding average temperature, traversing the temperature field model in turn All coordinate variables in , obtain a series of average temperatures, and use the least squares polynomial fitting method to obtain the temperature trend curve .

[0091] Furthermore, compare the temperature trend curve Trend curve with standard temperature Obtaining the exception segment number includes the following specific steps:

[0092] Determined based on the 3D model Functional segments in The range of segment coordinate variables in the axis direction, where The range of segment coordinate variables of each functional segment is , and Respectively The starting coordinate variables and the ending coordinate variables of each functional segment;

[0093] Calculate temperature trend curve Trend curve with standard temperature In the Temperature mean square error of each functional segment , No. Temperature mean square error of each functional segment Temperature trend curve Trend curve with standard temperature The square of the error in Segment coordinate variable range in the axis direction The definite integral within is divided by the length of the segment coordinate variable , the specific formula is as follows:

[0094] ;

[0095] in, express Coordinate variables in the axis direction, if Temperature mean square error of each functional segment If it is greater than or equal to the temperature mean square error threshold, the Functional segment abnormality, Marked as abnormal segment number, if Temperature mean square error of each functional segment If the error is less than the temperature mean square error threshold, the There is no abnormality in the functional segment.

[0096] like Figure 3 As shown, further, the gas concentration set Gas standard concentration set To obtain the exception segment number, the following specific steps are included:

[0097] Get gas concentration set Gas standard concentration set Middle Gas concentration vector of each functional segment and the gas standard concentration vector ;

[0098] Calculate the gas concentration vector and the gas standard concentration vector The mean square error of the concentration , as follows:

[0099] ;

[0100] in, For the Concentration error vector of each functional segment;

[0101] Jordi The mean square error of the concentration of each functional segment If it is greater than or equal to the concentration mean square error threshold, then the Functional segment abnormality, Marked as abnormal segment number, if The mean square error of the concentration of each functional segment If the concentration is less than the mean square error threshold, then the There is no abnormality in the functional segment.

[0102] Specifically, assuming the exception segment number is , No. Temperature difference of each functional segment Temperature trend curve Trend curve with standard temperature The error in Segment coordinate variable range in the axis direction The definite integral within is divided by the length of the segment coordinate variable , the specific formula is as follows:

[0103] ;

[0104] No. The concentration error vector of the functional segment , 、 、 and Respectively The oxygen concentration error, carbon dioxide concentration error, carbon monoxide concentration error and nitrogen concentration error of each functional segment.

[0105] Specifically, the fuzzy PID control algorithm includes fuzzy reasoning. The input variables of fuzzy reasoning include the temperature difference of the abnormal functional section, oxygen concentration error, carbon dioxide concentration error, carbon monoxide concentration error and nitrogen concentration error. The output variables of fuzzy reasoning include the proportional factor, integral factor and differential factor of the PID control algorithm. For each input variable or output variable, the corresponding domain is determined. The domain is the value range corresponding to the input variable or output variable. The fuzzy sets are defined as "negative large", "negative small", "zero", "positive medium" and "positive large". Fuzzy rules are formulated based on experience. The fuzzy rules establish the corresponding relationship between the input fuzzy set group and the output fuzzy set group. Among them, the input fuzzy set group is composed of the fuzzy sets to which the 5 input variables belong. Since there are 5 types of input variables, each input variable can belong to 5 different fuzzy sets. The input fuzzy set group has a total of 5 The output fuzzy set group is composed of the fuzzy sets to which the three output variables belong. The output fuzzy set group has a total of For each fuzzy rule, the membership of the five input variables to the corresponding fuzzy set in the input fuzzy set group is calculated by the Gaussian membership function, and the activation strength of the fuzzy rule is obtained by the AND operation. The activation strengths of all fuzzy rules are ANDed to determine the final output fuzzy set group and the membership of three output variables. The centroid method is combined with the domain of the three output variables for defuzzification, and the proportional factor, integral factor and differential factor of the PID control algorithm are output and further assigned to the PID control algorithm.

[0106] Specifically, the fuzzy PID control algorithm includes a PID control algorithm, which multiplies the temperature difference of the abnormal functional section, the oxygen concentration error, the carbon dioxide concentration error, the carbon monoxide concentration error and the nitrogen concentration error by the corresponding weight factors and sums them to generate a comprehensive error, obtains the proportional factor, integral factor and differential factor determined by fuzzy reasoning, and the proportional adjustment uses the product of the comprehensive error and the proportional factor as the proportional control amount. The proportional adjustment is used to reduce the comprehensive error at a macro level at one time to generate the residual comprehensive error. The integral adjustment integrates the residual comprehensive error and multiplies it by the integral factor to generate the integral control amount and update the residual comprehensive error. The integral adjustment is repeated. The integral adjustment is performed multiple times until the remaining comprehensive error is less than the comprehensive error threshold. The multiple integral adjustments are used to continuously reduce the remaining comprehensive error from a microscopic perspective. The differential adjustment multiplies the comprehensive error change rate of two adjacent integral adjustments by the differential factor to generate a differential correction control amount to avoid overshoot in the PID control algorithm. The number of differential adjustments is the same as the number of integral adjustments. The control amount of the PID control algorithm is equal to the sum of the proportional control amount, multiple integral control amounts and multiple differential control amounts. Considering the stoichiometric ratio of fuel and air, the control amount is split into the fuel valve adjustment amount and the air valve adjustment amount, and combined into the adjustment vector corresponding to the abnormal functional segment.

[0107] Example 2

[0108] like Figure 4 As shown, the present invention also discloses a decision control method based on temperature field and atmosphere analysis. The decision control system based on temperature field and atmosphere analysis is implemented, including the following specific steps:

[0109] At each monitoring moment, the temperature set is obtained by monitoring the segmented temperature measurement equipment group and gas detection equipment. and gas concentration sets ;

[0110] based on The geometric parameters of each functional segment are used to construct a three-dimensional model of the functional device and establish a Cartesian space coordinate system. is the total number of functional segments;

[0111] The variation function is used to describe the spatial variation characteristics of temperature in functional equipment, and the temperature set is The measured temperature in the 3D model is spatially expanded on the surface to generate an expanded temperature set. ;

[0112] The mesh segmentation technique is used to simplify and discretize the three-dimensional model into nodes, expand the temperature set through nearest neighbor search The temperature of each point in the graph is assigned to the corresponding surface node whose coordinates match the point. is the total number of nodes;

[0113] Get the ambient temperature , based on the thermal conductivity of functional devices Set up the heat conduction equation and define the convection boundary , the finite element matrix equations are constructed by the finite element method and the conjugate gradient method is used to solve the temperature field model ;

[0114] Temperature field model exist Statistical analysis is performed in the axial direction to generate temperature trend curves And with the standard temperature trend curve Compare and combine the gas concentration Gas standard concentration set The comparison result determines the abnormal segment number;

[0115] The temperature difference and concentration error vector of each abnormal segment number are used as the input of the fuzzy PID control algorithm. The fuzzy logic is used to dynamically adjust the PID control algorithm parameters, and the adjustment vector corresponding to each abnormal segment number is output to automatically adjust the fuel valve opening and air valve opening of the corresponding functional segment.

[0116] The present invention discloses a decision control system and method based on temperature field and atmosphere analysis, wherein the decision control system comprises a monitoring module, a trend analysis module, a decision module and an execution module; the monitoring module obtains a temperature set and a gas concentration set; the trend analysis module constructs a three-dimensional model and establishes a coordinate system, the variation function combines with the Kriging equation to supplement the temperature set to generate an extended temperature set, the discrete three-dimensional model is a plurality of nodes and is matched with points in the extended temperature set through nearest neighbor search, the heat conduction equation and the convection boundary are set and a finite element matrix equation group is constructed, the temperature field model is obtained by solving and a temperature trend curve is statistically generated; the decision module compares the temperature trend curve, the gas concentration set and the process standard to obtain the abnormal segment number, and adopts a fuzzy PID control algorithm to analyze the segment temperature difference and the concentration error vector of the abnormal segment number to generate an adjustment vector; the execution module automatically adjusts based on the adjustment vector, thereby realizing intelligent decision adjustment based on the joint analysis of the temperature field and the atmosphere.

[0117] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiment. All technical solutions based on the concept of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A decision-making control system based on temperature field and atmosphere analysis, characterized in that: Includes trend analysis module and decision-making module; The trend analysis module constructs a three-dimensional model and establishes a coordinate system, obtains a temperature set, uses a variation function to describe the spatial variation of temperature, adds temperatures of different points to the temperature set through the Kriging equation to generate an expanded temperature set, and uses a grid segmentation technique to simplify and discretize the three-dimensional model into Nodes, assign the temperature of each point in the expanded temperature set to the matching node through nearest neighbor search, set the heat conduction equation based on thermal conductivity and define the convection boundary in combination with the ambient temperature, use the finite element method to construct the finite element matrix equation group, solve the generated temperature field model through the conjugate gradient method and obtain the temperature trend curve by fitting. The temperature field model includes The temperature of the node, is the total number of nodes; The decision module obtains the temperature trend curve and gas concentration set and compares them with the process standards to determine the abnormal segment number, inputs the segment temperature difference and concentration error vector of each abnormal segment number into the fuzzy PID control algorithm, uses fuzzy logic to adjust the PID control algorithm parameters, and outputs the adjustment vector corresponding to each abnormal segment number, wherein the adjustment vector includes the fuel valve adjustment amount and the air valve adjustment amount.

2. The decision-making control system based on temperature field and atmosphere analysis according to claim 1, characterized in that: The step of adding the temperatures of different points to the temperature set by the Kriging equation to generate the expanded temperature set includes: Uniformly select multiple points to be estimated on the surface of the three-dimensional model; For each point to be estimated, a Kriging equation system is established based on the unbiasedness and minimum variance estimation principles to obtain an optimal weight coefficient vector, wherein the inner product of the optimal weight coefficient vector and the sequence of variance function values ​​of the coordinates of a single temperature measuring device in the temperature set is equal to the variance function value of the coordinates of the point to be estimated and the coordinates of the single temperature measuring device, and the sum of the optimal weight coefficients in the optimal weight coefficient vector is 1; Solve to obtain the optimal weight coefficient vector of the point to be estimated. The estimated temperature of the point to be estimated is the temperature concentration The inner product of the known temperature vector arranged by the measured temperatures and the optimal weight coefficient vector; Add the temperatures of all points to be estimated into the temperature set to generate an expanded temperature set.

3. The decision-making control system based on temperature field and atmosphere analysis according to claim 1, characterized in that: The use of a variation function to describe the spatial variation of temperature includes: From the temperature set Randomly select two temperature measuring device coordinates from the temperature measuring device coordinates to form a temperature measuring device pair, and calculate the spatial distance. is the total number of temperature measuring devices; The variogram value of the temperature measuring device pair is defined as the temperature semi-variance of the coordinates of the two temperature measuring devices, which is used to describe the temperature variability of spatial distance. The variogram is obtained by fitting the spatial distance and variogram value of all temperature measuring device pairs in the temperature set using the least squares method.

4. The decision-making control system based on temperature field and atmosphere analysis according to claim 1, characterized in that: Setting the heat conduction equation based on thermal conductivity and defining the convection boundary in combination with the ambient temperature includes: Determine thermal conductivity and set up the heat conduction equation to describe that the ratio of heat flux density to temperature gradient is equal to the inverse of thermal conductivity; The convection boundary is defined based on the convection heat transfer between the surface of the functional equipment and the ambient air. The heat flux density of the convection heat transfer is described by Newton's law of cooling, which is equal to the difference between the surface temperature and the ambient temperature multiplied by the convection heat transfer coefficient.

5. The decision-making control system based on temperature field and atmosphere analysis according to claim 1, characterized in that: The method of constructing a finite element matrix equation group by using the finite element method and solving the temperature field model by the conjugate gradient method includes: The heat conduction equation of a single unit is analyzed based on Fourier's law, and the weighted residual method is used to construct the finite element equation of a single unit. The finite element equation describes that the integral of the second-order temperature gradient in a single unit over the unit volume is proportional to the integral of the heat flux density at the unit boundary over the area of ​​the unit boundary, and the ratio is equal to the thermal conductivity; The temperature representation of any spatial coordinate in a single unit is described by interpolating the shape functions of all nodes in the unit. The shape functions are determined based on the geometric characteristics and node distribution of the unit. Substituting the temperature representation of a single unit into the finite element equation of the single unit, and organizing it into a finite element matrix equation through integration operation and matrix transformation, the finite element matrix equation describes that the product of the stiffness matrix of the single unit and the node temperature vector is equal to the load vector, the stiffness matrix is ​​used to describe the mutual influence between the temperatures of the nodes in the single unit, the node temperature vector records the temperature of all nodes in the single unit, and the load vector is used to describe the influence of the boundary heat flow on each node in the single unit; The finite element matrix equations of all units are updated and integrated into a finite element matrix equation group according to the assembly rules. The conjugate gradient method is used to obtain the node temperature vectors of all units and establish a temperature field model. The assembly rules are established based on the node sharing relationship of adjacent units. The values ​​of the shared nodes in the stiffness matrices and load vectors of adjacent units are accumulated and replaced respectively.

6. The decision-making control system based on temperature field and atmosphere analysis according to claim 1, characterized in that: The acquisition of the temperature trend curve and the gas concentration set and comparison with the process standards to determine the abnormal section number includes: Determine each functional segment based on the three-dimensional model The segment coordinate variable range in the axial direction is used to obtain the process standard, wherein the process standard includes a standard temperature trend curve and a gas standard concentration set; Calculate the temperature mean square error between the temperature trend curve and the standard temperature trend curve in a single functional segment. The temperature mean square error of a single functional segment is the definite integral of the square of the error between the temperature trend curve and the standard temperature trend curve within the corresponding segment coordinate variable range divided by the segment coordinate variable length. If the temperature mean square error is greater than or equal to the temperature mean square error threshold, the segment number of the single functional segment is marked as an abnormal segment number. If the temperature mean square error is less than the temperature mean square error threshold, no processing is performed. Extracting the gas concentration vector and gas standard concentration vector of a single functional segment in the gas concentration set and the gas standard concentration set; Calculate the concentration mean square error of a single functional segment. The concentration mean square error is the square of the two-norm of the concentration error vector of a single functional segment. , the concentration error vector is equal to the absolute value of the gas concentration vector of a single functional segment minus the gas standard concentration vector; If the concentration mean square error is greater than or equal to the concentration mean square error threshold, the segment number of the single functional segment is marked as an abnormal segment number. If the concentration mean square error is less than the concentration mean square error threshold, no processing is performed.

7. The decision-making control system based on temperature field and atmosphere analysis according to claim 1, characterized in that: The grid segmentation technology uses multiple straight lines parallel to the three coordinate axes to segment the three-dimensional model, defines the intersection of the straight lines as nodes, and defines the empty hexahedron surrounded by 8 nodes as a unit. The empty hexahedron refers to a hexahedron without any nodes inside. A nearest neighbor search is performed based on the coordinates of each point in the expanded temperature set to determine the node with the smallest spatial distance to the point coordinate and assign the temperature corresponding to the point coordinate.

8. The decision-making control system based on temperature field and atmosphere analysis according to claim 1, characterized in that: The fuzzy PID control algorithm includes fuzzy reasoning, and the input variables of the fuzzy reasoning include the temperature difference of the abnormal functional segment and the oxygen concentration error, carbon dioxide concentration error, carbon monoxide concentration error, and nitrogen concentration error in the concentration error vector. The output variables of the fuzzy reasoning include the proportional factor, integral factor, and differential factor of the PID control algorithm. The domain of each input variable or output variable is determined, the fuzzy set is defined, and fuzzy rules are formulated. The fuzzy rules establish the corresponding relationship between the input fuzzy set group and the output fuzzy set group. For a single fuzzy rule, the membership of each input variable to the corresponding fuzzy set in the input fuzzy set group is calculated by a Gaussian membership function, and the activation strength of the single fuzzy rule is obtained by an AND operation. The activation strengths of all fuzzy rules are ANDed to determine the membership of the output fuzzy set group and each output variable. The center of gravity method is used for defuzzification, and the proportional factor, integral factor, and differential factor are output.

9. The decision-making control system based on temperature field and atmosphere analysis according to claim 1, characterized in that: The decision control system further includes a monitoring module and an execution module; The monitoring module periodically sets monitoring moments and collects temperature sets and gas concentration sets at each monitoring moment, wherein the temperature sets and gas concentration sets respectively include The temperature sequence of each functional segment and The gas concentration vector of each functional segment, the temperature sequence of a single functional segment includes the measured temperatures of all temperature measuring devices in the single functional segment, is the total number of functional segments; The execution module automatically adjusts based on the adjustment vector of each abnormal segment number.

10. The decision-making control method based on temperature field and atmosphere analysis is characterized in that: The specific steps include: Obtain temperature sets and gas concentration sets at each monitoring moment; Construct a three-dimensional model based on the geometric parameters of each functional segment and establish a coordinate system; The spatial variation of temperature is described by using the variation function, and the temperature in the temperature set is spatially expanded based on the Kriging equation to generate an expanded temperature set; The structured grid technology is used to simplify and discretize the three-dimensional model into nodes, assign the temperature of each point in the expanded temperature set to the matching node through nearest neighbor search. is the total number of nodes; The heat conduction equation is set based on thermal conductivity and the convection boundary is defined by combining the obtained ambient temperature. The finite element matrix equation group is constructed by the finite element method and the conjugate gradient method is used to solve it to obtain the temperature field model. The temperature field model Statistical analysis is performed in the axial direction to fit and generate a temperature trend curve. The temperature trend curve and gas concentration set are compared with the corresponding process standards to determine the abnormal section number; The segment temperature difference and concentration error vector of each abnormal segment number are input into the fuzzy PID control algorithm, and the PID control algorithm parameters are adjusted using fuzzy logic. The adjustment vector corresponding to each abnormal segment number is output and automatically adjusted.

Images