Cooling control system

Abstract

The present invention discloses a cooling control system, which relates to the field of intelligent manufacturing and includes a time-varying analysis module, a monitoring module and an abnormality adjustment module; the time-varying analysis module considers heat conduction and water cooling to construct a regional temperature time-varying model of a tube mold sleeve, adopts the finite element method and the Galerkin method to discretely generate unit matrix equations, and constructs a regional temperature time-varying equation group through the time-discrete unit matrix equations; the monitoring module obtains all vertex temperatures and identifies abnormalities; the abnormality adjustment module performs multiple rounds of pre-control in the event of an abnormality, simulates the effect of each round of regulating valve opening based on the regional temperature time-varying equation group to decide whether to implement control or continue the next round of pre-control, and starts optimization pre-control if multiple rounds of pre-control are ineffective, simulates the effect of each round of joint control of cooling power, pumping power and valve opening by a rat swarm optimization algorithm based on the regional temperature time-varying equation group to obtain the optimal solution and implement it, thereby realizing centrifugal casting cooling control that considers regional temperature interaction and control cost.

Description

Technical Field

[0001] The present invention relates to the field of intelligent manufacturing, and in particular to a cooling control system. Background Art

[0002] In centrifugal casting, cooling control is a key factor in determining casting quality. Scientific and precise cooling control can achieve uniform and refined internal structure in castings, effectively avoiding defects such as shrinkage, cracks, and loose structure caused by uneven cooling. This significantly improves the mechanical properties and dimensional accuracy of castings, while also reducing scrap rates and shortening production cycles, effectively increasing production efficiency, reducing production costs, and ensuring process stability.

[0003] Existing cooling control technologies for centrifugal casting generally suffer from the defect of extensive zoning control and fail to fully consider the impact of temperature interactions between different areas, resulting in large temperature control deviations and significant fluctuations in cooling effects. This cannot meet the requirements of modern centrifugal casting for high-precision and dynamic cooling control. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the present invention proposes a cooling control system to realize a high-precision centrifugal casting cooling control system that synergistically considers regional temperature interaction and control costs.

[0005] The technical solutions for achieving the purpose of the present invention are:

[0006] A cooling control system includes a time-varying analysis module, a monitoring module, and an abnormality adjustment module;

[0007] The time-varying analysis module divides the tube mold into For each region, a regional temperature time-varying model is constructed based on the physical properties of the tube mold sleeve, the law of heat conduction, and the characteristics of cooling water flow. The finite element method is used to discretize the region into units and determine the unit temperature expression through the volume function. The Galerkin method is used to integrate the regional temperature time-varying model on each unit and combine it with the unit temperature expression to derive the unit matrix equation. The unit matrix equations of all units in the region are combined and the regional temperature time-varying equation group is constructed through time discretization. is the total number of regions;

[0008] The monitoring module obtains a three-dimensional thermal map through a stereo matching algorithm, combined with The temperature of the sampling points in the region is fitted with a temperature determination function, the vertex coordinates in the region are determined based on the finite element unit discrete method, and the vertex temperature is calculated based on the temperature determination function, which is consistent with the target temperature. Compare to identify temperature anomalies;

[0009] The abnormal adjustment module performs multiple rounds of pre-regulation when an abnormality occurs, obtaining All vertex temperatures in the region, the valve opening vector is updated using the PID algorithm in each round of pre-control And through the regional temperature time-varying equations to simulate and update the first index of this round, based on the convergence of the first index, the decision is based on the valve opening vector Implement control or continue the next round of pre-control until the first indicator of the maximum number of rounds has not converged, then start the optimization pre-control and reduce the cooling power , pumping power and valve opening vector Integrate into solution vector , the optimization problem is constructed with minimizing the second index as the optimization goal. In each round of iteration, the solution vector of each mouse is updated by the mouse swarm optimization algorithm and substituted into the regional temperature time-varying equation group to simulate and update the corresponding second index until the second index converges to obtain the optimal solution vector And implement regulation, the first indicator is the weighted sum of the total mean square error of temperature and the total junction temperature gradient, and the second indicator is equal to the sum of the first indicator and the regulation cost.

[0010] Furthermore, the regional temperature time-varying model is constructed based on Fourier's heat conduction law. Specifically, the regional time-varying term is equal to the heat conduction term minus the heat exchange term. The time-varying term represents the rate of change of regional temperature over time. The heat conduction term reflects the effect of the heat conduction of the tube mold sleeve on the regional temperature change. The heat conduction term is equal to the thermal conductivity coefficient of the tube. The heat exchange term reflects the effect of heat exchange between the tube mold sleeve and the cooling water on the regional temperature change, and is proportional to the regional heat exchange area and the regional heat exchange coefficient, and is proportional to the specific heat capacity of the pipe. , tube density Inversely proportional to the area volume and the cooling water temperature There is a linear relationship.

[0011] Specifically, the finite element method discretizes each region into a finite number of tetrahedral units, each of which includes 4 vertices. The unit temperature expression is constructed based on the small-scale approximation principle. The point temperature of any point in the unit can be obtained by linear interpolation of the vertex temperatures of the 4 vertices in the unit combined with the corresponding volume function, where the volume function of a single vertex is defined as the volume ratio of the tetrahedron formed by any point in the unit and the three vertices other than the single vertex and the tetrahedron unit.

[0012] Specifically, the Galerkin method is used to take the volume function of each vertex in the unit temperature expression as the weight function and perform volume integration on the regional temperature time-varying model in the unit to offset the residual of the partial differential equation solution. The time-varying term, heat conduction term and heat exchange term in the regional temperature time-varying model and the integral of the volume function of each vertex in the unit are calculated and sorted to generate the unit matrix equation. The unit matrix equation is specifically the unit heat conduction matrix and the element vertex temperature vector The product of the element heat capacity matrix and the element vertex temperature vector About Time The product of the partial differential terms of is equal to the element load vector , where the unit heat conduction matrix No. Rank The value of the column reflects the The temperature change of the first vertex is affected by heat conduction The influence of temperature change of each vertex, the unit vertex temperature vector Including the vertex temperature of the 4 vertices in the unit, the unit heat capacity matrix No. Rank The value of the column reflects the The temperature change of the vertex The effect of stored heat on the vertices, the element load vector Middle The value of the row reflects the heat exchange for the The external influence of the temperature change of each vertex, .

[0013] Specifically, all unit matrix equations in a single region are combined to form a unit matrix equation group for a single region, specifically the product of the regional heat conduction matrix and the regional vertex temperature matrix of a single region plus the regional heat capacity matrix and the regional vertex temperature matrix with respect to time. The product of the partial differential terms is equal to the regional load matrix. The implicit difference format is used to time discretize the partial differential terms and replace the partial differential terms in the unit matrix equations. The implicit difference format will be the interval length The partial differential term in the region is approximately the temperature matrix of the regional vertices in the interval The temperature change within the interval is divided by the interval length , multiply both sides of the equation by the interval duration Arrange and construct a set of time-varying equations for regional temperature.

[0014] Furthermore, the monitoring module includes a temperature sensing unit, an infrared imaging unit and an abnormality judgment unit;

[0015] The temperature sensing unit is The temperature sensors in each area measure the temperature of the sampling points in each area respectively;

[0016] The infrared imaging unit uses infrared imaging devices with two viewing angles to obtain the left thermal map and the right thermal map. Based on the stereo matching algorithm, the left thermal map and the right thermal map are matched and spliced ​​into a three-dimensional thermal map.

[0017] Abnormal judgment unit basis The coordinates corresponding to the temperature of the sampling points in each area are combined with the brightness of the corresponding points in the three-dimensional thermal map. The least squares method is used to fit the temperature determination function. The coordinates of all vertices in each area are obtained by unit discretization based on the finite element method, and the brightness of the corresponding points in the three-dimensional thermal map are extracted. The corresponding vertex temperature is determined according to the temperature determination function, and the temperature anomaly is judged based on whether there is an area with a temperature mean square error greater than or equal to the warning error.

[0018] Furthermore, in each region’s temperature time-varying equations, except for the cooling water temperature All parameters except the regional heat exchange coefficient are fixed as a priori conditions, and the cooling water temperature The regional heat exchange coefficient needs to be modeled as a variable. The coefficient determination function reflecting the correlation between the regional heat exchange coefficient of a single area and the regional water flow is determined based on the Ditus-Bolt formula. The regional water flow of a single area is equal to the valve opening of the electric control valve of the single area and the starting water flow provided by the water pump. The product of the starting water flow The pumping power of the water pump The associated flow determination function is obtained by least square fitting, reflecting the cooling water temperature Cooling power of heat exchange equipment Associated water temperature determination function It is obtained by fitting through the least squares algorithm. In multiple rounds of pre-control, it is only necessary to substitute the equation relationship between the regional water flow and the valve opening of each area into the coefficient determination function, and replace the regional heat exchange coefficient in the regional temperature time-varying equation group for simulation. In the optimization pre-control, it is necessary to substitute the product of the flow determination function of each area and the corresponding valve opening into the coefficient determination function and replace the regional heat exchange coefficient in the regional temperature time-varying equation group, and simultaneously replace the cooling water temperature in the regional temperature time-varying equation group with the water temperature determination function for simulation.

[0019] Furthermore, multiple rounds of pre-regulation include the following specific steps:

[0020] Will All vertex temperatures in the region are taken as the vertex temperatures of round 0, and the parameter values ​​except the regional heat exchange coefficient in the regional temperature time-varying equations are obtained and fixed according to the prior conditions and variable modeling. The electronically controlled valve opening of each region generates the valve opening vector of round 0 , set the maximum number of rounds ;

[0021] Calculate the number of The mean square error of temperature in each region , based on the The boundary line and search distance between each region and each adjacent region Determine the left and right boundary vertices, and The boundary temperature difference between a region and each adjacent region and the search distance The average ratio of The boundary temperature gradient of the 0th round of the region ;

[0022] A new round of simulation begins. The mean square error of temperature and the boundary temperature gradient for each zone in the previous round are input into the PID algorithm to obtain the valve opening adjustment vector for this round. This vector is summed with the valve opening vector updated in the previous round and constrained according to the valve opening range. The valve opening vector for this round is updated and converted into the regional heat exchange coefficient vector for this round based on variable modeling.

[0023] Based on interval length Substitute the regional heat exchange coefficient of each region in the regional heat exchange coefficient vector into the corresponding regional temperature time-varying equation group, and substitute the vertex temperature of the 0th round for simulation, calculate the first index after this round of simulation and determine whether it is less than the first minimum value;

[0024] If it is less than the first minimum value, control is performed based on the valve opening vector of the current round; if it is greater than or equal to the first minimum value, whether the number of rounds in the current round is equal to the maximum number of rounds is determined;

[0025] If it is less than the maximum number of rounds, it will jump to a new round of simulation. If the number of rounds in this round is equal to the maximum number of rounds, it will start optimization pre-control.

[0026] Furthermore, optimizing pre-regulation includes the following specific steps:

[0027] Will All vertex temperatures in the region are taken as vertex temperatures in round 0, and the temperature time-varying equations of the fixed region are fixed according to the prior conditions except for the cooling water temperature. and parameter values ​​other than the regional heat exchange coefficient, integrating the cooling power , pumping power and valve opening vector Constructing the solution vector ;

[0028] The optimization problem is constructed with the goal of minimizing the second indicator, which is equal to the first indicator plus the control cost, and the control cost is equal to the cooling power. , pumping power The sum of the valve opening penalty term and the valve opening penalty term is used to constrain The opening of the electronically controlled valve in each area is within the range of 0 to 1, which is equal to the penalty factor multiplied by The sum of the inverse rectangular window functions of the regions;

[0029] Randomly initialize in the solution vector space mice, each mouse's position is a set of random solution vectors, is the total number of mice;

[0030] The position of each mouse in round 0 is converted into the cooling water temperature and regional heat exchange coefficient vector of each mouse in round 0 based on variable modeling, combined with the interval length , the vertex temperature in round 0 and the regional temperature time-varying equations of each area are simulated, the second index of each mouse in round 0 is calculated and the position of the mouse with the smallest second index in round 0 is taken as the optimal position in round 0;

[0031] A new round of simulation begins. The mouse at the optimal position in the previous round becomes the leader mouse of this round, and the remaining mice become servant mice of this round. The position of each servant mouse in this round is updated to the position of the previous round plus the forward direction of this round multiplied by the random step length of this round. The forward direction of each servant mouse is the optimal position of the previous round minus the position of each servant mouse in the previous round. The random step length of each servant mouse is equal to a random number between 0 and 1 multiplied by the exploration step length inversely proportional to the number of iterations.

[0032] The position of each mouse in this round is converted into the cooling water temperature and regional heat exchange coefficient vector in this round based on the variable modeling, combined with the interval length , the vertex temperature of round 0 and the regional temperature time-varying equations of each area are simulated, the second index of each mouse in this round is calculated and the position of the mouse with the smallest second index in this round is taken as the optimal position of this round;

[0033] Determine whether the smallest second index in this round is less than the second minimum value. If it is less than the second minimum value, the optimal position of this round is used as the optimal solution vector. And implement regulation, if it is greater than or equal to the second minimum value, jump to start a new round of simulation.

[0034] The controlled device includes a centrifugal casting pipe mold sleeve, a motor, a conduit, a water pump, a heat exchange device, a water tank, an electric control valve and a temperature sensor; the centrifugal casting pipe mold sleeve is used for centrifugal casting; the motor drives the centrifugal casting pipe mold sleeve to rotate at high speed to generate centrifugal force; the conduit is used to transport cooling water; the water pump is used to extract cooling water and inject it into the conduit; the heat exchange device cools the recovered cooling water and injects it into the water tank; the water tank is used to store cooling water; the electric control valve is remotely controlled by the cooling control system to control the regional water flow entering each area; the temperature sensor is used to collect the sampling point temperature of each area.

[0035] Compared with the prior art, the present invention partitions the tube mold sleeve and constructs a regional temperature time-varying model considering the heat conduction and water cooling characteristics, adopts the finite element method and the Galerkin method to discretely generate unit matrix equations, combines the unit matrix equations and constructs a regional temperature time-varying equation group through time discretization. When temperature abnormality is detected, multiple rounds of pre-control are first performed, and the effect of each round of valve opening control is simulated based on the regional temperature time-varying equation group to decide whether to implement control or continue the next round of pre-control. If multiple rounds of pre-control are ineffective, the optimization pre-control is started, and the rat swarm optimization algorithm is simulated based on the regional temperature time-varying equation group to jointly control the cooling power, pumping power and valve opening in each round to obtain the optimal solution and implement it, thereby realizing high-precision centrifugal casting cooling control that takes into account regional temperature interaction and control cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 Schematic diagram of the cooling control system of the present invention;

[0037] Figure 2 Schematic diagram of the volume function in the present invention;

[0038] Figure 3 This is a flowchart of multiple rounds of pre-regulation in the present invention;

[0039] Figure 4 Schematic diagram of the controlled device in the present invention.

[0040] Reference numerals: 10, centrifugal casting pipe mold sleeve; 20, motor; 30, conduit; 40, water pump; 50, heat exchange equipment; 60, water tank; 70, electric control valve; 80, temperature sensor. DETAILED DESCRIPTION

[0041] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0042] Example 1:

[0043] like Figure 1 As shown, the present invention discloses a cooling control system, including a time-varying analysis module, a monitoring module and an abnormality adjustment module;

[0044] The time-varying analysis module divides the tube mold sleeve into equal volumes along the axial direction of the tube mold sleeve. A regional temperature time-varying model is constructed based on the physical properties of the tube mold sleeve, the law of heat conduction and the cooling water flow characteristics. The finite element method is used to discretize the region into a finite number of units at equal intervals and determine the corresponding unit temperature expression through the volume function of each vertex in the unit. The Galerkin method is used to integrate the regional temperature time-varying model on each unit and combine it with the corresponding unit temperature expression to obtain the unit matrix equation. The unit matrix equations of all units in the region are combined and the regional temperature time-varying equation group is constructed through time discretization. The regional temperature time-varying equation group reflects the cooling water temperature. and water flow The influence of the change on the diffusion of regional temperature, is the total number of regions;

[0045] The monitoring module obtains the temperature of the sampling points in each area through multiple temperature sensors, and splices the thermal images taken by infrared imaging devices at different viewing angles into a three-dimensional thermal map of the pipe mold sleeve based on the stereo matching algorithm. The temperature of the sampling point in each area is obtained by fitting the brightness of the sampling point coordinates in the three-dimensional thermal map to obtain the temperature determination function. The vertex coordinates of all units in each area are determined according to the unit discretization method of the finite element method to obtain the corresponding vertex temperature and compare it with the target temperature. Compare to determine whether there are abnormal areas;

[0046] The abnormal adjustment module performs multiple rounds of pre-adjustment when there are abnormal areas to obtain The vertex temperature of all units in the region, in each round of pre-control, the PID algorithm is used to update the valve opening vector based on the first indicator of the previous round And update the first index of this round through the regional temperature time-varying equations simulation. If the first index converges, it is based on the valve opening vector Regulation If the first indicator of the maximum number of rounds has not converged, the optimization pre-control will be started to reduce the cooling power of the heat exchange equipment. , Pumping power of the pump and valve opening vector Integrate into solution vector , the optimization problem is constructed with minimizing the second index as the optimization goal. In each round of iteration, the solution vector of each mouse is updated by the mouse swarm optimization algorithm and substituted into the regional temperature time-varying equation group to simulate and update the corresponding second index. When the second index converges, the optimal solution vector is obtained. and regulate heat exchange equipment, water pumps and The electronically controlled valve in each area, where the valve opening vector include The valve opening of the electric control valve in each area, the first indicator is based on The second index is determined based on the temperature mean square error of each area and the boundary temperature gradient of adjacent areas, and the second index is determined based on the first index and the control cost.

[0047] Furthermore, the regional temperature time-varying model is constructed based on Fourier's heat conduction law. The change of the temperature of the tube mold sleeve over time depends on the heat conduction of the tube mold sleeve and the heat exchange between the tube mold sleeve and the cooling water. The regional temperature time-varying model of each region follows the time-varying term equal to the heat conduction term minus the heat exchange term, as follows:

[0048] ;

[0049] Among them, the left side is a time-varying term, indicating the Regional temperature of the area Over time The rate of change, ;

[0050] One item on the right is the heat conduction term, reflecting the effect of the heat conducted by the tube mold sleeve on the regional temperature The impact of changes in and are the thermal conductivity coefficient of the tube mold material and the regional temperature Laplace operator, Laplace operator Specifically, the regional temperature The sum of the second-order gradients in the three coordinate axis directions of the spatial rectangular coordinate system. Heat conduction refers to the spontaneous transfer of heat from high temperature to low temperature, thereby causing temperature changes in the region and indirectly affecting the regional temperature of adjacent regions;

[0051] Right Binomial It is the heat exchange item, which reflects the effect of heat exchange between the tube mold sleeve and cooling water on the regional temperature. The impact of the changes in Regional heat exchange area and regional heat transfer coefficient Directly proportional to the specific heat capacity of the tube mold material , tube density of tube mold sleeve Hedi The volume of the region Inversely proportional to the cooling water temperature Shows a linear relationship, where the regional heat exchange area The regional heat exchange coefficient is determined by the layout of the cooling water pipes. With the Regional water flow in the region Positively correlated.

[0052] Specifically, such as Figure 2 As shown in the figure, the finite element method discretizes each area in the tube mold into a finite number of tetrahedral units. Each tetrahedral unit includes 4 vertices. Based on the small-scale approximation principle, when the unit geometry is small, the point temperature of any point in the unit obeys a linear distribution in space, that is, the point temperature of any point in the unit It can be obtained by linear interpolation of the vertex temperatures of the four vertices in the unit combined with the corresponding volume function, and described by the unit temperature expression, as follows:

[0053] ;

[0054] in, is the coordinate of any point in the cell, 、 and are the spatial rectangular coordinate systems Axis value, Axis value and Axis value, and The units are The volume function and vertex temperature corresponding to the vertex, Vertices ( Figure 2 The volume function of the black dot Defined as any point inside the cell ( Figure 2 The middle shaded point) and The three vertices other than the vertex ( Figure 2 Hollow center point) and the volume ratio of the tetrahedron and tetrahedron unit, when confirming the coordinates of any point When Volume function of vertices The value of the volume function of the first vertex The details are as follows:

[0055] ;

[0056] in, 、 and The first The vertices in the rectangular coordinate system of space Axis value, Axis value and Axis value, determinant value is equal to 6 times the volume of the tetrahedron formed by the points corresponding to the four point coordinates, .

[0057] Specifically, the regional temperature time-varying model is in the form of a partial differential equation. It is difficult to accurately satisfy the partial differential equation, which will produce a residual. The Galerkin method uses the volume function of each vertex in the unit temperature expression as a weight function, and performs volume integration on the regional temperature time-varying model within the unit to offset the residual in a weighted manner, so that the residual of the regional temperature time-varying model is 0 from the perspective of the unit as a whole. Then Volume function of vertices The volume integral of the regional temperature time-varying model as a weight function is as follows:

[0058] ;

[0059] in, is the temperature at any point in the unit, is the unit heat exchange coefficient, which is equal to the regional heat exchange coefficient of the area where the unit is located. and The unit volume and unit heat exchange area are equal to the regional volume and regional heat exchange area of ​​the unit divided by the total number of units in the region. Since the regions and units are divided into equal volumes and discrete intervals, the unit volume of any unit in all regions of the tube mold sleeve is equal to and unit heat exchange area are all the same, ;

[0060] The temperature inside the unit Replace it with the unit temperature expression, and calculate the time-varying term, heat conduction term and heat exchange term in the regional temperature time-varying model and the Volume function of vertices The points are as follows:

[0061] ;

[0062] ;

[0063] ;

[0064] in, For the unit Volume function of vertices Hamiltonian operator, Hamiltonian operator Specifically, the volume function The sum of the first-order gradients in the three coordinate axis directions of the spatial rectangular coordinate system, For the unit The vertex temperature corresponding to each vertex is ;

[0065] The volume integral of the regional temperature time-varying model when the four vertices in the unit are used as weight functions is obtained respectively and the unit matrix equation is generated. The unit matrix equation is as follows:

[0066] ;

[0067] in, is the element vertex temperature vector, 、 and They are the unit heat conduction matrix, unit heat capacity matrix and unit load vector, and their dimensions are 、 and , unit heat conduction matrix No. Rank Heat transfer elements of the column , unit heat capacity matrix No. Rank Heat capacity elements of the column and the element load vector No. Load element of the row The details are as follows:

[0068] ;

[0069] ;

[0070] ;

[0071] Among them, the heat conduction element Reflects the unit The temperature change of the first vertex is affected by heat conduction The internal effect caused by the temperature change of each vertex, the heat capacity element Reflects the unit The temperature change of the vertex The effect of stored heat on the vertices, load elements Reflecting the heat exchange The external effects caused by the temperature changes of each vertex are all present in the integral calculation results.

[0072] Furthermore, the construction of the regional temperature time-varying equations includes the following specific steps;

[0073] Combination No. All the unit matrix equations in the region constitute the The unit matrix equations of the region are as follows:

[0074] ;

[0075] in, For the The first temperature vector generated by concatenating all the cell vertex temperature vectors in the region The regional vertex temperature matrix of the region, For the The first one generated by splicing the heat conduction matrices of all units in the region The regional heat conduction matrix of the region, For the The first one generated by splicing the heat capacity matrices of all units in the region The regional heat capacity matrix of the region, For the The first one generated by concatenating all the element load vectors in the region The regional load matrix of the region, ;

[0076] Because the The regional vertex temperature matrix of the region Changes over time and can be replaced with Time-dependent variables , using implicit difference format to Partial differential terms in the unit matrix equations of the region For time discretization, the implicit difference format shows that The regional vertex temperature matrix of the region from Time to Partial differential term of time Can be approximated as Regional vertex temperature matrix at time minus Regional vertex temperature matrix at time The resulting temperature change is divided by the interval duration. , as follows:

[0077] ;

[0078] The first The regional vertex temperature matrix in the element matrix equation system of the region Replace with Regional vertex temperature matrix at time , using the implicit difference format to replace Partial differential term of time , multiply both sides of the equation by the interval duration Remove the fractional structure and organize the generated regional temperature time-varying equations, specifically: .

[0079] Furthermore, the monitoring module includes a temperature sensing unit, an infrared imaging unit and an abnormality judgment unit;

[0080] The temperature sensing unit is deployed in the tube mold sleeve The temperature sensors in each area measure the temperature of the sampling points in each area respectively. The coordinates of the sampling points in each area are the same as the installation coordinates of the corresponding temperature sensors.

[0081] The infrared imaging unit uses infrared imaging devices arranged on both sides of the tube mold sleeve to simultaneously capture the left and right thermal maps of the tube mold sleeve. Based on the stereo matching algorithm, the left and right thermal maps are matched and spliced ​​to generate a three-dimensional thermal map of the tube mold sleeve. The three-dimensional thermal map reflects the temperature of each coordinate of the tube mold sleeve. The stereo matching algorithm is an existing algorithm and will not be elaborated on in detail.

[0082] The abnormality judgment unit will be based on the installation coordinates of the temperature sensor The sampling point temperatures of each area are placed on the corresponding points of the three-dimensional thermal map and the brightness of the corresponding points is extracted. The temperature determination function is obtained by least square fitting. The unit discretization method based on the finite element method is used to obtain the coordinates of all vertices of all units in each area and extract the brightness corresponding to the vertex coordinates in the three-dimensional thermal map. The corresponding vertex temperature is determined according to the temperature determination function, and the difference between the vertex temperature in each area and the target temperature is statistically calculated. If there is an area with a temperature mean square error greater than or equal to the warning error, it is considered to be an abnormal area.

[0083] Specifically, the material, size, area division method of the pipe mold sleeve and the layout of the cooling water pipe in each area have been determined as prior conditions. Thermal conductivity of tubes in the time-varying equations of regional temperature in each region , tube density Specific heat capacity of pipe , regional volume and regional heat exchange area Is a fixed value, the interval length Equal to the acceptable adjustment time of the tube mold sleeve, only the cooling water temperature and regional heat transfer coefficient Can be regarded as variables. Multiple rounds of pre-regulation and optimized pre-regulation require variable modeling to build the relationship between variables and adjustment quantities. Adjustment quantities include the cooling power of heat exchange equipment. , Pumping power of the pump and valve opening vector , .

[0084] Furthermore, the variable modeling of multiple rounds of pre-regulation and optimized pre-regulation includes the following specific steps:

[0085] No. Regional heat transfer coefficient of each region With the Regional water flow in the region is positively correlated, then The coefficients of the region determine the function It is determined based on the Ditus-Bolt formula, as follows:

[0086] ;

[0087] in, 、 、 and are the thermal conductivity, specific heat, viscosity and density of cooling water, respectively. and are the inner diameter and cross-sectional area of ​​the catheter, respectively. ;

[0088] No. Regional water flow in the region Equal to Valve opening of the electronically controlled valve in each zone and the starting water flow provided by the pump The product of , valve opening Between 0 and 1;

[0089] Start water flow The pumping power of the water pump Directly related, pre-fitting the flow determination function by the least squares method ;

[0090] Cooling water temperature Cooling power of heat exchange equipment Directly related, pre-fitting the water temperature determination function through the least squares algorithm ;

[0091] The adjustment amount in multiple rounds of pre-control is only the valve opening vector , pumping power and cooling power Unchanged, cooling water temperature With starting water flow Considered as a constant, for area, and the regional water flow and valve opening Substitute the equation relationship into the coefficient to determine the function And replace the regional heat exchange coefficient in the regional temperature time-varying equations You can simulate;

[0092] Optimize the adjustment quantity in pre-control including the cooling power of heat exchange equipment , Pumping power of the pump and valve opening vector , for the area, the flow determination function Multiply by valve opening Then substitute the coefficients to determine the function And replace the regional heat exchange coefficient in the regional temperature time-varying equations , and simultaneously determine the water temperature function Replace the cooling water temperature in the time-varying regional temperature equations You can perform simulation.

[0093] like Figure 3 As shown, further, multiple rounds of pre-regulation include the following specific steps:

[0094] Will The vertex temperatures of all units in the region are taken as the vertex temperatures of the 0th round, and the parameter values ​​except the regional heat exchange coefficient in the regional temperature time-varying equations are obtained and fixed according to the prior conditions and variable modeling. The electronically controlled valve opening of each region generates the valve opening vector of round 0 , set the maximum number of rounds ;

[0095] Calculate the number of The mean square error of temperature in each region , as follows:

[0096] ;

[0097] in, and The target temperature and Region No. The vertex temperature of each vertex in round 0, , is the total number of vertices in a single region, ;

[0098] Get the The boundary line between the area and each adjacent area, and the left and right sides of the boundary line whose distance to the boundary line is less than or equal to the search distance The vertices of the left and right boundaries are regarded as the left and right boundary vertices, and the junction temperature difference is calculated as the average vertex temperature of the left boundary vertices minus the average vertex temperature of the right boundary vertices. The boundary temperature difference between the region and each adjacent region is divided by the search distance The mean of The boundary temperature gradient of the 0th round of the region , among which, if or , then There is only one adjacent area in each region, and in the rest of the cases Each region has two adjacent regions in the axial direction;

[0099] Start a new round of simulation, and use the temperature mean square error and boundary temperature gradient of each area in the previous round as the input of the PID algorithm. The PID algorithm outputs the valve opening adjustment vector of this round through the proportional term, integral term and differential term. The valve opening adjustment vector includes Valve opening adjustment amount of the electronically controlled valve in each area;

[0100] The valve opening adjustment vector of this round is summed with the valve opening vector updated in the previous round to obtain the preliminary valve opening vector of this round. Based on the range of valve openings, the valve openings greater than 1 and less than 0 in the preliminary valve opening vector are adjusted to 1 and 0 respectively. The valve opening vector of this round is updated and converted into the regional heat exchange coefficient vector of this round based on variable modeling.

[0101] Based on interval length Substitute the regional heat exchange coefficient of each region in the regional heat exchange coefficient vector into the corresponding regional temperature time-varying equations, and substitute the vertex temperature of the 0th round for deduction and simulation. After the simulation, calculate the temperature mean square error and boundary temperature gradient of each region in this round.

[0102] Sum up the current round The temperature mean square error and the junction temperature gradient of each region are obtained to obtain the total temperature mean square error and the total junction temperature gradient of this round, and weighted summed by a preset weight ratio to obtain the first index of this round, and determine whether the first index of this round is less than the first minimum value;

[0103] If it is less than the first minimum value, then the valve opening vector based on this round is If the electric control valve in each area is greater than or equal to the first minimum value, it is further determined whether the number of rounds in this round is equal to the maximum number of rounds;

[0104] If the number of rounds in this round is less than the maximum number of rounds, a new round of simulation will be started. If the number of rounds in this round is equal to the maximum number of rounds, it means that multiple rounds of pre-control are invalid and optimized pre-control is started.

[0105] Furthermore, optimizing pre-regulation includes the following specific steps:

[0106] Will The vertex temperatures of all units in the region are taken as the vertex temperatures of the 0th round, and the regional temperature time-varying equations except the cooling water temperature are obtained and fixed according to the prior conditions. and the parameter values ​​other than the regional heat exchange coefficient, define the solution vector ,in, 、 and are the cooling power of the heat exchange equipment, the pumping power of the water pump, and the valve opening vector respectively;

[0107] Construct an unconstrained optimization problem to minimize the second indicator, that is, minimize the first indicator plus the control cost, where the first indicator is used to evaluate the temperature control effect and the control cost is used to minimize the cooling power. Add pumping power Simultaneous constraints The opening of the electric control valve in each area is within the range of 0 to 1, and the control cost is equal to the cooling power Add pumping power Add a penalty term for valve opening, where the penalty term for valve opening is equal to the penalty factor multiplied by The sum of the inverse rectangular window functions of the regions, the penalty factor is a predetermined very large positive number, Valve opening of each zone When the valve opening is within the range of 0 to 1, the corresponding inverse rectangular window function value is 0. When it is less than 0 or greater than 1, the corresponding inverse rectangular window function value is 1. The product of the sum of the inverse rectangular window functions of the regions and the penalty factor ensures that the update of the valve opening in the rat swarm optimization algorithm does not exceed the range. In this embodiment, the penalty factor is set to 10000.

[0108] Randomly initialize in the solution vector space mice, the position of each mouse is a set of random solution vectors, is the total number of mice;

[0109] The first The position of the rat in round 0 was transformed into the The cooling water temperature and regional heat exchange coefficient vector of the mouse in round 0, according to the interval length Substitute the cooling water temperature and the regional heat exchange coefficient of each region in the regional heat exchange coefficient vector into the corresponding regional temperature time-varying equations, and perform simulations based on the vertex temperature of the 0th round to calculate the The second index of the mouse in round 0 is the position of the mouse with the smallest second index in round 0 as the optimal position in round 0. ;

[0110] Start a new round of simulation, and take the mouse with the best position obtained in the previous round as the leader mouse of this round, and the rest Each mouse is used as a servant mouse in this round. The position of each servant mouse in this round is updated by the position of the previous round plus the forward direction of this round multiplied by the random step length of this round. The forward direction of each servant mouse in this round is the optimal position of the previous round minus the position of each servant mouse in the previous round. The random step length of each servant mouse is equal to a random number between 0 and 1 multiplied by the exploration step length inversely proportional to the number of iteration rounds.

[0111] The first The position of the mouse in this round is converted into the first The cooling water temperature and regional heat exchange coefficient vector of each mouse in this round, according to the interval length Substitute the cooling water temperature and the regional heat exchange coefficient of each region in the regional heat exchange coefficient vector into the corresponding regional temperature time-varying equation group, and substitute the vertex temperature of the 0th round for deduction and simulation to calculate the The second index of the mice in this round, the position of the mouse with the smallest second index in this round is regarded as the optimal position of this round;

[0112] Determine whether the smallest second index in this round is less than the second minimum value. If it is less than the second minimum value, the optimal position of this round is used as the optimal solution vector. And heat exchange equipment, water pumps and The electronically controlled valves in each area are regulated. If it is greater than or equal to the second minimum value, a new round of simulation is started.

[0113] Example 2:

[0114] like Figure 4 As shown, the present invention also discloses a controlled device, including a centrifugal casting pipe mold sleeve 10, a motor 20, a conduit 30, a water pump 40, a heat exchange device 50, a water tank 60, an electric control valve 70 and a temperature sensor 80;

[0115] The centrifugal casting pipe mold sleeve 10 is used for centrifugal casting. Figure 4 The middle dashed line represents the regional boundary;

[0116] The motor 20 drives the centrifugal casting pipe mold sleeve 10 to rotate at high speed to generate centrifugal force. Figure 4 The encircled arrow in the middle represents the rotation direction of the motor 20;

[0117] The conduit 30 is used to transport cooling water to achieve zone cooling of the centrifugal casting pipe mold sleeve 10. Figure 4 The continuous triangle in the middle represents the cooling water delivery direction;

[0118] The water pump 40 draws cooling water from the water tank 60 and injects it into the conduit 30;

[0119] The heat exchange device 50 cools the cooling water recovered through the conduit 30 and injects the cooling water into the water tank 60;

[0120] The water tank 60 is used to store the cooling water after the heat exchange device 50 cools down;

[0121] The electric control valve 70 is respectively arranged in each area of ​​the centrifugal casting pipe mold sleeve 10, and is remotely controlled by the cooling control system to control the regional water flow entering the corresponding area by adjusting the valve opening;

[0122] The temperature sensors 80 are respectively arranged in each area of ​​the centrifugal casting pipe mold sleeve 10 to collect the temperature of the sampling point in each area.

[0123] The present invention discloses a cooling control system, comprising a time-varying analysis module, a monitoring module and an abnormality adjustment module; the time-varying analysis module constructs a regional temperature time-varying model of a pipe mold sleeve by considering heat conduction and water cooling characteristics, adopts the finite element method and the Galerkin method to discretely generate unit matrix equations, combines the unit matrix equations of the same region and constructs a regional temperature time-varying equation group through time discretization; the monitoring module obtains all vertex temperatures and identifies abnormalities; the abnormality adjustment module performs multiple rounds of pre-control in the event of an abnormality, simulates the effect of each round of regulating valve opening based on the regional temperature time-varying equation group to decide whether to implement control or continue the next round of pre-control, and starts optimized pre-control if multiple rounds of pre-control are ineffective, simulates the effect of each round of joint regulation of cooling power, pumping power and valve opening by the rat swarm optimization algorithm based on the regional temperature time-varying equation group to obtain the optimal solution and implement it, thereby realizing centrifugal casting cooling control considering regional temperature interaction and control cost.

[0124] 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 cooling control system, characterized in that: It includes time-varying analysis module, monitoring module and abnormal adjustment module; The time-varying analysis module divides the tube mold sleeve into The regional temperature time-varying model is constructed by using the finite element method to discretize the region into units. The volume integral of the regional temperature time-varying model in each unit is discretized into a unit matrix equation through the volume function using the Galerkin method. The unit matrix equations of all units in the region are combined and the regional temperature time-varying equation group is constructed based on time discretization. is the total number of regions; The monitoring module is based on the unit discrete method of the finite element method, combined with the collected sampling point temperature and three-dimensional thermal map, to obtain All vertex temperatures in a region are compared with the target temperature to identify abnormalities; The abnormal adjustment module performs multiple rounds of pre-control when the temperature of the centrifugal casting of the pipe mold sleeve is abnormal. In each round of pre-control, the valve opening vector is updated by the PID algorithm, and the regional temperature time-varying equation group is used to simulate and update the first indicator and decide whether to implement control or continue the next round of pre-control. If multiple rounds of pre-control are invalid, the optimized pre-control is started, and the cooling power, pumping power and valve opening vector are integrated into the solution vector. The solution vector of each mouse is updated in each round of iteration through the mouse swarm optimization algorithm and the second indicator is updated in combination with the regional temperature time-varying equation group. The optimal solution vector is iteratively obtained and control is implemented. The first indicator is the weighted sum of the total mean square error of temperature and the total junction temperature gradient, and the second indicator is equal to the first indicator plus the control cost.

2. The cooling control system according to claim 1, wherein: The regional temperature time-varying model is constructed based on Fourier's heat conduction law, that is, the regional time-varying term is equal to the heat conduction term minus the heat exchange term. The time-varying term represents the rate of change of the regional temperature over time. The heat conduction term reflects the influence of the heat conduction of the tube mold sleeve on the regional temperature change. The heat conduction term is equal to the product of the tube thermal conductivity coefficient and the Laplace operator of the regional temperature. The heat exchange term reflects the influence of the heat exchange between the tube mold sleeve and the cooling water on the regional temperature change. It is proportional to the regional heat exchange area and the regional heat exchange coefficient, inversely proportional to the specific heat capacity of the tube, the tube density and the regional volume, and linearly related to the cooling water temperature.

3. The cooling control system according to claim 1, wherein: The finite element method discretizes each region through tetrahedral units, where the tetrahedral units include four vertices. A unit temperature expression is constructed based on the principle of small-scale approximation. The point temperature of any point within the tetrahedral unit is obtained by linear interpolation of the vertex temperatures of the four vertices within the unit combined with the corresponding volume function, wherein the volume function of a single vertex is defined as the volume ratio of the tetrahedron formed by any point within the unit and the three vertices other than the single vertex to the tetrahedron unit.

4. The cooling control system according to claim 1, wherein: The Galerkin method performs volume integration on the regional temperature time-varying model within the unit and calculates the integral of the time-varying term, the heat conduction term, the heat exchange term and the volume function of each vertex in the unit respectively, and generates a unit matrix equation. The unit matrix equation includes a unit heat conduction matrix, a unit heat capacity matrix and a unit load vector, wherein the unit heat conduction matrix is ​​used to describe the influence of the temperature change of a vertex in the unit on the temperature change of other vertices through heat conduction, the unit heat capacity matrix is ​​used to describe the influence of the temperature change of a vertex in the unit on the stored heat of other vertices, and the unit load vector is used to describe the external influence of heat exchange on the temperature change of the vertices in the unit.

5. The cooling control system according to claim 1, wherein: The construction of the regional temperature time-varying equation group includes: All unit matrix equations in the combined region constitute the unit matrix equation group of the region. The implicit difference format is used to replace the partial differential terms in the unit matrix equation group with the corresponding time discrete representation, and then it is organized into a regional temperature time-varying equation group. The implicit difference format expresses the partial differential terms within the interval time as the temperature change of the regional vertex temperature matrix within the interval time divided by the interval time.

6. The cooling control system according to claim 1, wherein: The multiple rounds of pre-regulation and optimized pre-regulation require the determination of prior conditions and variable modeling, including: The parameters of the regional temperature time-varying equations except the cooling water temperature and the regional heat exchange coefficient are taken as prior conditions and fixed; The cooling water temperature is modeled as a variable, and a water temperature determination function reflecting the relationship between cooling water temperature and cooling power is obtained through least squares fitting. Variable modeling was performed on the regional heat exchange coefficient. Based on the Dittus-Bolt formula, a coefficient determination function reflecting the relationship between the regional heat exchange coefficient and the regional water flow was determined. The regional water flow was equal to the product of the regional valve opening and the starting water flow. The flow determination function reflecting the relationship between the starting water flow and the pumping power was obtained through least squares fitting. Multiple rounds of pre-control are simulated by substituting the equations of regional water flow and valve opening into the coefficient determination function and replacing the regional heat exchange coefficient in the regional temperature time-varying equations; The optimized pre-control substitutes the product of the flow determination function and the valve opening into the coefficient determination function, and replaces the regional heat exchange coefficient and cooling water temperature in the regional temperature time-varying equation group with the substituted coefficient determination function and water temperature determination function respectively for simulation.

7. The cooling control system according to claim 6, wherein: The multiple rounds of pre-regulation include: Will The vertex temperatures of all regions in the 0th round are taken as the vertex temperatures. According to the prior conditions and variable modeling, the parameter values ​​except the regional heat exchange coefficient are obtained. The valve opening vector of the 0th round is obtained and the maximum number of rounds is set. The mean square error of the temperature and the boundary temperature gradient of each region in the 0th round are calculated. A new round of simulation begins. The mean square error of temperature and the boundary temperature gradient of each region in the previous round are input into the PID algorithm to obtain the valve opening adjustment vector for this round. The updated valve opening vector from the previous round is superimposed and constrained based on the valve opening range. The valve opening vector for this round is updated and converted into the regional heat exchange coefficient vector for this round based on variable modeling. Based on the regional heat exchange coefficient vector, the vertex temperature of the 0th round and the time-varying equations of all regional temperatures, simulation is performed within the interval time. The first indicator after this round of simulation is calculated and whether it converges is determined. If it converges, regulation is implemented based on the valve opening vector of this round. If it does not converge, it is determined whether the maximum number of rounds has been reached. If not, a new round of simulation is started. If it has been reached, optimization pre-regulation is started.

8. The cooling control system according to claim 6, wherein: The optimized pre-regulation includes: Will All vertex temperatures in the region are taken as the vertex temperatures of round 0. According to the prior conditions, the parameters other than the cooling water temperature and the regional heat exchange coefficient are obtained. The cooling power, pumping power and valve opening vector are integrated to construct the solution vector. The optimization problem is constructed with the goal of minimizing the second index, which is equal to the sum of the first index, cooling power, pumping power and valve opening penalty. The valve opening penalty is equal to the penalty factor multiplied by The sum of the inverse rectangular window functions of the regions; Random initialization mice, the position of each mouse is a set of solution vectors, is the total number of mice; The position of each mouse in round 0 is converted into the cooling water temperature and regional heat exchange coefficient vector in round 0 based on variable modeling. The simulation is combined with the interval length, the vertex temperature in round 0, and the regional temperature time-varying equation group to calculate the second index of each mouse in round 0 and determine the optimal position in round 0. A new round of simulation begins. Rats that were not in the optimal position in the previous round are randomly explored in the direction of the optimal position in the previous round based on their positions in the previous round to update their positions in this round. The position of each rat in this round is converted into the cooling water temperature and regional heat exchange coefficient vector of this round based on variable modeling. The simulation is combined with the interval length, the vertex temperature in round 0, and the regional temperature time-varying equations to calculate the second indicator of each rat in this round and determine the optimal position of this round. Determine whether the second indicator corresponding to the optimal position of this round is less than the second minimum value. If it is less than the second minimum value, the optimal position of this round is used as the optimal solution vector and regulation is implemented. If it is greater than or equal to the second minimum value, jump to start a new round of simulation.

9. The cooling control system according to claim 1, wherein: The monitoring module includes a temperature sensing unit, an infrared imaging unit and an abnormality judgment unit; The temperature sensing unit is The temperature sensors in each area measure the temperature of the sampling points in each area respectively; The infrared imaging unit uses infrared imaging devices with two viewing angles to obtain the left thermal map and the right thermal map. Based on the stereo matching algorithm, the left thermal map and the right thermal map are matched and spliced ​​into a three-dimensional thermal map. Abnormal judgment unit basis The coordinates corresponding to the temperature of the sampling points in each area are combined with the brightness of the corresponding points in the three-dimensional thermal map. The least squares method is used to fit the temperature determination function. The coordinates of all vertices in each area are obtained by unit discretization based on the finite element method, and the brightness of the corresponding points in the three-dimensional thermal map are extracted. The corresponding vertex temperature is determined according to the temperature determination function, and the temperature anomaly is judged based on whether there is an area with a temperature mean square error greater than or equal to the warning error.

Images