Temperature field analysis method of heat exchange module

Abstract

The present invention relates to the technical field of heat exchange modules, and discloses a temperature field analysis method for a heat exchange module. The method comprises: performing multi-source data acquisition on a microchannel dual-phase precision phase change heat exchange module to obtain temperature field acquisition data, pressure data, flow data and phase change material characteristic data; performing dimensionless number analysis and calculation to obtain flow type criterion values, and calculating gas-liquid two-phase flow heat transfer coefficient distribution data; performing coupling calculation to obtain phase change region heat transfer characteristic data; performing conjugate gradient iteration calculation to obtain temperature field reconstruction data; inputting the temperature field reconstruction data into a temperature field evolution prediction model to perform temperature field evolution prediction to obtain temperature field prediction data; establishing a dual-objective optimization function including heat exchange efficiency and energy consumption according to the temperature field prediction data, and solving to obtain heat exchanger control parameters. The present invention improves the accuracy of heat exchange performance prediction, and at the same time, obtains the optimal control parameters that take into account heat exchange efficiency and energy consumption.

Description

Technical Field

[0001] The present invention relates to the technical field of heat exchange modules, and in particular to a temperature field analysis method of a heat exchange module. Background Art

[0002] Microchannel heat exchange technology has shown great advantages in the field of high heat flux density heat dissipation due to its large specific surface area and high heat transfer coefficient. However, the complex flow pattern inside the microchannel and the variable gas-liquid two-phase flow state make it difficult to predict the heat exchange performance, which restricts the optimal design and control of the microchannel heat exchanger.

[0003] Phase change energy storage technology can significantly improve the energy density and temperature control accuracy of heat exchange systems, but the heat transfer mechanism in the phase change process is complex, involving the coupling of multiple heat transfer modes such as heat conduction, convection heat transfer and phase change latent heat, and the temperature field distribution in the phase change area is difficult to accurately measure and predict. Most of the existing temperature field analysis methods are based on simplified models, ignoring key influencing factors such as flow morphology and phase change characteristics, resulting in a large deviation between the analysis results and actual working conditions. Summary of the invention

[0004] The present invention provides a temperature field analysis method for a heat exchange module, which improves the accuracy of heat exchange performance prediction and obtains optimal control parameters that take into account both heat exchange efficiency and energy consumption.

[0005] In a first aspect, the present invention provides a temperature field analysis method for a heat exchange module, the temperature field analysis method for a heat exchange module comprising:

[0006] Perform multi-source data collection on the microchannel dual-phase precision phase change heat transfer module to obtain temperature field collection data, pressure data, flow data and phase change material characteristic data;

[0007] Performing dimensionless number analysis and calculation on the pressure data and the flow data to obtain flow pattern criterion values, and calculating gas-liquid two-phase flow heat transfer coefficient distribution data according to the flow pattern criterion values;

[0008] Substituting the gas-liquid two-phase flow heat transfer coefficient distribution data and the phase change material characteristic data into the phase change heat transfer control equation for coupled calculation to obtain the phase change region heat transfer characteristic data;

[0009] According to the heat transfer characteristic data of the phase change region and the temperature field acquisition data, a radial basis function interpolation algorithm is input to perform conjugate gradient iterative calculation to obtain temperature field reconstruction data;

[0010] Inputting the temperature field reconstruction data into a temperature field evolution prediction model to perform temperature field evolution prediction to obtain temperature field prediction data;

[0011] A dual-objective optimization function including heat exchange efficiency and energy consumption is established according to the temperature field prediction data, and the heat exchanger control parameters are obtained by iterative solution through a particle swarm algorithm.

[0012] In a second aspect, the present invention provides a temperature field analysis device for a heat exchange module, the temperature field analysis device for a heat exchange module comprising:

[0013] The acquisition module is used to collect multi-source data of the microchannel dual-phase precision phase change heat transfer module to obtain temperature field acquisition data, pressure data, flow data and phase change material characteristic data;

[0014] An analysis module, used for performing dimensionless number analysis and calculation on the pressure data and the flow data to obtain a flow pattern criterion value, and calculating gas-liquid two-phase flow heat transfer coefficient distribution data according to the flow pattern criterion value;

[0015] A calculation module, used for substituting the gas-liquid two-phase flow heat transfer coefficient distribution data and the phase change material characteristic data into the phase change heat transfer control equation for coupling calculation to obtain the phase change region heat transfer characteristic data;

[0016] An iteration module, used for inputting a radial basis function interpolation algorithm to perform conjugate gradient iterative calculation according to the heat transfer characteristic data of the phase change region and the temperature field acquisition data, so as to obtain temperature field reconstruction data;

[0017] A prediction module, used for inputting the temperature field reconstruction data into a temperature field evolution prediction model to perform temperature field evolution prediction and obtain temperature field prediction data;

[0018] The solution module is used to establish a dual-objective optimization function including heat exchange efficiency and energy consumption according to the temperature field prediction data, and obtain the heat exchanger control parameters through iterative solution of the particle swarm algorithm.

[0019] The third aspect of the present invention provides a temperature field analysis device for a heat exchange module, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory to enable the temperature field analysis device for the heat exchange module to execute the above-mentioned temperature field analysis method for the heat exchange module.

[0020] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the temperature field analysis method of the heat exchange module.

[0021] In the technical solution provided by the present invention, a dimensionless analysis model considering flow characteristics is established, combined with the distribution data of gas-liquid two-phase flow heat transfer coefficient, the flow heat transfer characteristics inside the microchannel are accurately described, and the accuracy of heat transfer performance prediction is improved; a temperature field reconstruction method based on multi-source data fusion is adopted, and temperature sensor data and phase change region heat transfer characteristic data are comprehensively utilized to achieve high-precision reconstruction of the heat exchanger temperature field; a deep learning prediction model with physical constraints is introduced, and the physical rationality of the temperature field prediction results is guaranteed through energy conservation constraints and attention mechanisms; a multi-objective optimization method based on an adaptive particle swarm algorithm is constructed, and the optimal control parameters that take into account both heat transfer efficiency and energy consumption are obtained through the entropy weight method and TOPSIS decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0023] Figure 1 A schematic diagram of a flow chart of a temperature field analysis method for a heat exchange module provided in an embodiment of the present application;

[0024] Figure 2 A schematic block diagram of the structure of a temperature field analysis device for a heat exchange module provided in an embodiment of the present application;

[0025] Figure 3 A schematic block diagram of the structure of a temperature field analysis device for a heat exchange module provided in an embodiment of the present application. DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change based on actual conditions.

[0028] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0029] It should be further understood that the term “and / or” used in the specification and appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0030] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0031] See also Figure 1 , Figure 1 A schematic diagram of a flow chart of a temperature field analysis method for a heat exchange module provided in an embodiment of the present application, such as Figure 1 As shown, the temperature field analysis method of the heat exchange module provided in the embodiment of the present application includes steps S100 to S600.

[0032] Step S100, performing multi-source data acquisition on the microchannel dual-phase precision phase change heat transfer module to obtain temperature field acquisition data, pressure data, flow data and phase change material characteristic data;

[0033] It is understandable that the execution subject of the present invention may be a temperature field analysis device of a heat exchange module, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0034] Specifically, the sensor array of the microchannel two-phase precision phase change heat exchange module is subjected to data acquisition and processing, and the original data of the temperature field in the heat exchange process is monitored in real time through the sensor array arranged in the heat exchange module. The temperature sensor measurement signal is spatially interpolated according to the original data of the temperature field, and the temperature distribution is inferred between the measurement points by the mathematical interpolation method to obtain the temperature field acquisition data. At the same time, the pressure sensor data at the inlet and outlet of the microchannel two-phase precision phase change heat exchange module is sampled and processed, and the pressure data in the system is obtained by measuring the pressure difference between the inlet and outlet of the fluid, reflecting the flow resistance characteristics inside the heat exchange module. The mass flow meter data of the microchannel two-phase precision phase change heat exchange module is sampled and processed to obtain flow data, reflecting the rate at which the fluid flows through the heat exchange module. The temperature measurement data of the thermocouple array of the phase change energy storage chamber inside the microchannel two-phase precision phase change heat exchange module is used for phase change state identification and calculation. The phase change process is a key link in the heat exchange module, which directly determines the heat storage and heat release characteristics of the system. The temperature change in the phase change energy storage chamber is monitored in real time by a thermocouple array. Based on the temperature measurement data, the physical state of the phase change material is analyzed, and the phase change completion degree is identified and calculated to obtain the completion degree data of different stages in the phase change process. The phase change completion degree data is used to describe the progress of the current phase change process and can help analyze the release or absorption of latent heat in the phase change process. Based on the phase change completion degree data, the phase change process is characterized by analysis to determine the phase change start time and phase change duration. These two parameters are important characteristics in the phase change process and directly affect the heat transfer efficiency of the heat exchange module and the utilization efficiency of the phase change material. The phase change start time reflects the time point when the phase change material enters the phase change state from the solid or liquid state, while the phase change duration describes the length of time that the phase change material maintains in the phase change state. The phase change start time and phase change duration are substituted into the phase change energy storage calculation model to obtain the heat storage capacity data. The phase change energy storage calculation model describes the heat that the material can store or release during the phase change process through physical model simulation and mathematical calculation. These data help evaluate the heat storage capacity of the heat exchange module and determine its applicable working conditions. By calculating the heat storage capacity data, the energy storage and heat exchange capacity of the phase change material under specific working conditions can be effectively evaluated. According to the heat storage capacity data, the characteristics of the phase change material are calculated to obtain the characteristic data of the phase change material, including physical parameters such as phase change latent heat, thermal conductivity, specific heat capacity, etc., which reflect the changes in the thermal properties of the phase change material during the phase change process.

[0035] Step S200, performing dimensionless number analysis and calculation on the pressure data and the flow data to obtain a flow pattern criterion value, and calculating the gas-liquid two-phase flow heat transfer coefficient distribution data according to the flow pattern criterion value;

[0036] Specifically, the Reynolds number, Weber number and Kapitza number are calculated based on the pressure and flow data to obtain the dimensionless data of the flow characteristics in the microchannel. The Reynolds number is used to characterize the relative size between the inertial force and the viscous force of the fluid, reflecting whether the flow state is laminar or turbulent; the Weber number is used to describe the relationship between the inertial force and the surface tension of the fluid, and is used to judge the generation and splitting of droplets and bubbles; the Kapitza number reflects the mutual influence of gravity, viscosity and surface tension, especially in the phase change process, the judgment of the flow morphology is particularly important. These dimensionless numbers jointly characterize the flow characteristics. By making interval judgments on the dimensionless data of the flow characteristics in the microchannel, a preliminary judgment is made on the flow type. When the Weber number is less than 0.5, the flow form of the system is judged to be an intermittent slug flow region. In this region, the liquid phase and the gas phase are alternately formed, and the fluid presents a morphology similar to a bubble projectile. This flow type has a more significant bubble slip phenomenon, and its intermittent characteristics lead to the volatility of the heat transfer effect. If the Weber number is between 0.5 and 2.5, it is judged as an annular flow area. At this time, the liquid phase forms a thin film along the channel wall, while the gas phase flows in the central area. The characteristic of this annular flow is the fluctuation of the liquid film surface, which will significantly affect the heat transfer process and the pressure distribution of the system. When the Weber number is greater than 2.5, the flow of the system is judged as a mist flow area. At this time, the strong action of the gas phase causes the liquid phase to be split into a large number of droplets and dispersed in the gas phase, forming a mixture similar to a mist. The heat transfer characteristics of the flow type are affected by the entrainment behavior of the droplets. Therefore, the mist flow usually exhibits a higher convective heat transfer capacity. According to the flow type area judgment data, the flow type conversion characteristic parameters are calculated to obtain the flow type judgment value. After obtaining the flow type judgment value, the boundary layer thickness in the microchannel is calculated to obtain the boundary layer characteristic data. The thickness of the boundary layer directly affects the heat transfer between the fluid and the wall. Especially in the gas-liquid two-phase flow, the fluctuation and development of the boundary layer play an important role in the heat transfer performance. Based on the boundary layer characteristic data and flow pattern criterion values ​​in the microchannel, the bubble slip coefficient is calculated for the intermittent slug flow area to obtain the bubble slip characteristic data. The bubble slip coefficient is a dimensionless quantity that describes the movement speed of the gas phase in the liquid phase relative to the overall flow velocity, which helps to analyze the phase interface movement and the heat transfer effect between the gas and liquid phases in the intermittent slug flow. For the annular flow area, the liquid film fluctuation characteristics are calculated according to the flow pattern criterion values ​​to obtain the liquid film fluctuation characteristic data. The fluctuation of the liquid film not only affects the pressure drop of the fluid, but also has a significant impact on the heat transfer coefficient, because the fluctuation increases the disturbance of the interface and enhances the turbulent mixing, thereby improving the heat transfer effect. At the same time, for the mist flow area, the droplet entrainment coefficient is calculated according to the flow pattern criterion values ​​to obtain the droplet entrainment characteristic data. In the mist flow, the phenomenon of droplet entrainment is very common. The airflow drives a large number of droplets to move in the channel. The presence of such droplets greatly increases the heat exchange area and improves the heat transfer coefficient.However, the entrainment of droplets will also cause fluctuations in system pressure. Accurate description of the entrainment characteristics is helpful for analyzing the stability of the heat exchange system and the heat transfer effect. The bubble slip characteristic data, liquid film fluctuation characteristic data and droplet entrainment characteristic data are substituted into the two-phase flow heat transfer coefficient calculation model to obtain the heat transfer coefficient distribution data of gas-liquid two-phase flow.

[0037] Step S300, substituting the gas-liquid two-phase flow heat transfer coefficient distribution data and the phase change material characteristic data into the phase change heat transfer control equation for coupling calculation to obtain the phase change region heat transfer characteristic data;

[0038] Specifically, the heat storage capacity data in the phase change material characteristic data is piecewise linearized to obtain piecewise linearized phase change characteristic data, so that the entire phase change process can be considered linear in each small segment. The phase change region is gridded according to the gas-liquid two-phase flow heat transfer coefficient distribution data to generate grid data of the phase change region, and the entire phase change region is divided into multiple small calculation units to facilitate accurate description and calculation of the local characteristics of the phase change process. The piecewise linearized phase change characteristic data is substituted into the coupling model of the heat source heat removal cycle and the refrigeration cycle to obtain the cyclic coupling heat transfer equation. The coupling model aims to simultaneously consider the different thermodynamic behaviors of the phase change material in the heat source heat removal process and the refrigeration process, and describe them uniformly in one heat transfer equation. On the basis of the cyclic coupling heat transfer equation, the natural convection term is modified and calculated to better reflect the contribution of natural convection in the heat exchange process. In this way, the phase change heat transfer control equation is established, and its expression is: In this equation, represents the density of the phase change material, is the specific heat capacity, is the temperature field, For time, is the velocity field, is the thermal conductivity, is the latent heat of phase change, represents the liquid fraction, is the volumetric thermal expansion coefficient, is the acceleration due to gravity, The phase change heat transfer control equation comprehensively considers the complex thermodynamic behaviors of phase change materials such as heat conduction, convection and phase change latent heat release or absorption during the heat transfer process, and can fully describe the temperature change and dynamic characteristics of the phase change region. When establishing the phase change heat transfer control equation, the phase change heat transfer control equation is discretized according to the grid data of the phase change region to obtain a discretized control equation group. In the discretization process, numerical methods such as finite difference method, finite volume method or finite element method are used to discretize the continuous control equation in space and time to obtain a set of equations suitable for numerical solution. The discretized equation group can concretize the thermodynamic state changes of each grid unit in the phase change region, so as to facilitate accurate numerical simulation of the entire region. Then, the discretized control equation group is solved by time advancement. By gradually advancing the time step, the temperature field data of the entire phase change region that changes with time is obtained, and the dynamic simulation of the phase change process is realized. The heat flux density distribution in the phase change region is calculated according to the temperature field data of the phase change region. Heat flux is an important parameter to measure the heat transfer rate, reflecting how heat is transferred inside the material and the intensity of the transfer during the phase change process. By calculating the heat flux distribution in the phase change region, more intuitive heat transfer characteristic data can be obtained. The temperature field data and heat flux distribution data of the phase change region are comprehensively processed, and the heat transfer behavior and laws of the entire phase change process are analyzed in combination with the changes in the temperature field and the distribution of heat flux, so as to obtain the heat transfer characteristic data of the phase change region.

[0039] Step S400, inputting the heat transfer characteristic data of the phase change region and the temperature field acquisition data into the radial basis function interpolation algorithm to perform conjugate gradient iterative calculation to obtain temperature field reconstruction data;

[0040] Specifically, a radial basis function matrix and a distance weight matrix are constructed according to the temperature field acquisition data and the radial basis function interpolation algorithm to obtain an initial interpolation matrix. The radial basis function interpolation algorithm is a multidimensional interpolation method. Through a set of sampling points distributed in space, a continuous interpolation function is constructed to achieve reconstruction of the temperature field in the entire space. The role of the distance weight matrix is ​​to describe the degree of mutual influence between different sampling points, and these influence weights are used to more accurately capture the changing characteristics of the temperature field. The heat transfer characteristic data of the phase change region is normalized to obtain normalized heat transfer characteristic data. The normalized heat transfer characteristic data is substituted into the initial interpolation matrix to obtain a corrected interpolation coefficient matrix. Based on the corrected interpolation coefficient matrix, the objective function of temperature field reconstruction is constructed. The expression of the temperature field reconstruction objective function is: ,in, Reconstruct the objective function for the temperature field, is the radial basis function interpolation matrix, is the temperature field to be reconstructed, is the measured temperature data, is the regularization parameter, is the Laplace operator. The objective function consists of two parts: the first part is the interpolation error, that is, the difference between the reconstructed temperature field and the measured temperature, which is used to minimize the deviation of the reconstructed temperature field; the second part is the regularization term, which smoothes the temperature field through the Laplace operator. is a regularization parameter that balances the minimization of errors and the smoothness of the temperature field, ensuring that the reconstructed temperature field is consistent with the measured data and not too complex to avoid overfitting. The temperature field reconstruction objective function is gradient calculated to obtain the gradient data of the objective function. The direction of change of the objective function in the parameter space is determined to find the direction that minimizes the objective function value. By solving the gradient of the objective function, the sensitivity information of the objective function to the temperature field is obtained, which in turn guides how to adjust the temperature field to minimize the objective function. Based on the gradient data of the objective function, the conjugate direction is updated to obtain the search direction data. The conjugate gradient method is an effective solution method for large linear systems. By selecting a search direction that is conjugate to the directions of the previous steps in each iteration, the convergence speed is significantly accelerated and the amount of calculation is reduced. In the conjugate direction update of each step, the new search direction is orthogonal to the previous direction in a sense, thereby ensuring that the value of the objective function can be effectively reduced in each iteration. The search direction data is substituted into the linear search algorithm to obtain the optimal step length data. The role of the linear search algorithm is to find an optimal step size in the current search direction so that the value of the objective function is minimized after moving in this direction. In this way, it is ensured that the steps of each iteration move in the optimal direction, making the reconstruction of the temperature field more efficient. Based on the optimal step size data, the temperature field is iteratively updated to gradually reconstruct the distribution of the entire temperature field. After each iteration, the residual is calculated, that is, the difference between the reconstructed temperature field and the measured temperature field. When the iterative residual is less than the specified threshold 10⁻ 6 When , the iteration is stopped, indicating that the accuracy of temperature field reconstruction has reached the predetermined requirement, and the final temperature field reconstruction data is obtained, reflecting the temperature distribution in the entire phase change region.

[0041] Step S500, inputting the temperature field reconstruction data into the temperature field evolution prediction model to perform temperature field evolution prediction to obtain temperature field prediction data;

[0042] Specifically, the temperature field reconstruction data is input into the encoding layer of the temperature field evolution prediction model for processing. The role of the encoding layer is to extract the feature information in the temperature field reconstruction data. In the encoding layer, there are three convolution sublayers. The size of the convolution kernel of the first convolution sublayer is 3×3, the step size is set to 1, and the number of output channels is 64. The purpose is to extract local features while maintaining spatial resolution. The convolution kernel of the second convolution sublayer is also 3×3, the step size is set to 2, and the number of output channels is increased to 128. By increasing the step size, the size of the feature map is gradually reduced to obtain higher-level features. The convolution kernel size of the third convolution sublayer is 3×3, the step size is 2, and the number of output channels reaches 256. Through layer-by-layer convolution, the input temperature field data is converted into a set of feature encoding data, which contains the main spatial information and feature distribution in the input temperature field. The temperature field feature encoding data is input into the long short-term memory network (LSTM) layer of the temperature field evolution prediction model for temporal feature extraction to capture the evolution law of the temperature field in the time dimension. The long short-term memory network layer contains four parallel memory units, each of which has an input gate, a forget gate, and an output gate. These gate structures work together to control which information needs to be retained and which information needs to be discarded when processing time series data. The number of neurons in the hidden layer is set to 512 to ensure that there is enough capacity to learn the complex time series characteristics of the temperature field. Through the processing of the LSTM network layer, the dynamic evolution information in the temperature field feature encoding data is extracted to obtain the time series feature data, which describes the change relationship between the temperature field at different time points and captures the long-term dependence and short-term volatility in the temperature field change. In order to ensure the accuracy of the temperature field evolution prediction, the energy conservation constraint calculation is performed on the time series feature data to obtain the constrained time series feature data. Energy conservation is an important physical constraint condition in the temperature field change. By performing energy conservation constraints on the time series feature data, it is ensured that the prediction results of the model conform to the laws of physics and improve the credibility and physical consistency of the prediction. The constrained time series feature data is input into the attention mechanism layer of the temperature field evolution prediction model for spatial correlation analysis. The attention mechanism layer adopts a self-attention structure, with 8 attention heads and 256 hidden layer dimensions. These settings allow the model to pay attention to the spatial correlation between different regions when processing temperature field data, especially in complex fluid heat exchange and phase change processes, where heat transfer between different regions has significant spatial correlation. The self-attention mechanism aggregates the features of different regions by weighted summation, thereby highlighting the influence of important regions and obtaining spatial attention feature data, which helps to better understand the interaction between various parts during the evolution of the temperature field. The spatial attention feature data is input into the decoding layer of the temperature field evolution prediction model to reconstruct the temperature field. The decoding layer is used to restore the feature data after previous encoding and time series processing to the original temperature field form and predict the temperature distribution at future time points.The decoding layer consists of three deconvolution sublayers. The convolution kernel size of the first deconvolution sublayer is 4×4, the step size is 2, and the number of output channels is 128. The deconvolution operation increases the spatial resolution of the feature map and gradually restores the shape of the original input. The convolution kernel size of the second deconvolution sublayer is 4×4, the step size is also 2, and the number of output channels is 64. Upsampling continues to expand the feature data to a size close to the original input size. The convolution kernel size of the third deconvolution sublayer is 3×3, the step size is 1, and the number of output channels is 1. The feature data is restored to the resolution and shape consistent with the original temperature field, and the final temperature field prediction data is obtained, reflecting the evolution of the temperature field in the future time period.

[0043] Step S600: establishing a dual-objective optimization function including heat exchange efficiency and energy consumption according to the temperature field prediction data, and obtaining heat exchanger control parameters through iterative solution using a particle swarm algorithm.

[0044] Specifically, the first optimization objective function and the second optimization objective function are constructed according to the temperature field prediction data, wherein the first optimization objective function represents the heat transfer efficiency, is the inlet temperature, is the outlet temperature, is the ambient reference temperature. The physical meaning of this objective function is to measure the temperature variation of the heat exchanger under different working conditions, thereby reflecting its heat transfer efficiency. The second optimization objective function is used to characterize energy consumption, where is the fluid density, is the volume flow rate, The function is used to evaluate the energy consumption of the entire system. The goal is to minimize the energy consumption of the system while meeting the heat exchange requirements. The two objective functions are combined by weighted summation to obtain the dual-objective optimization function. ,in and are the weight coefficients of heat transfer efficiency and energy consumption, respectively, and . The weighted summation method allows different weights to be assigned to heat transfer efficiency and energy consumption according to the needs of specific applications, so as to balance the relationship between the two under different working conditions and find the optimal control parameter combination. Particles are initialized based on the dual-objective optimization function. The first type of particles and the second type of particles are initialized to obtain the data of the initial particle swarm with a double-layer structure. The first type of particles are used to optimize the heat transfer efficiency, while the second type of particles are used to optimize the energy consumption. Through the particle swarm with a double-layer structure, the two optimization objectives are iteratively solved at the same time to achieve the optimal compromise between efficiency and energy consumption. The initialization of the particle swarm is completed by randomly generating the position and velocity of the initial particles, so that the search space can cover all possible solution areas. The generated data of the initial particle swarm with a double-layer structure is input into the hierarchical fitness calculation model for fitness calculation. The first layer of the hierarchical fitness calculation model is used to calculate the heat transfer efficiency. The fitness value of this layer reflects the contribution of each particle to improving the heat transfer efficiency; the second layer is used to calculate the energy consumption, and the fitness value represents the performance of each particle in reducing energy consumption. Through the hierarchical fitness calculation method, the multi-objective performance of particles is effectively evaluated to help identify the particles with the best performance. The two-layer fitness data is standardized by the minimum-maximum normalization method to obtain the comprehensive fitness data. Based on the comprehensive fitness data, the particles are non-dominated sorted to obtain the hierarchical optimal position data. Non-dominated sorting is an important tool for multi-objective optimization. By sorting particles according to their dominance relationship, particles that perform well on different objectives are identified. Based on the hierarchical optimal position data, the particle speed is adaptively updated to obtain adaptive particle speed data. The adaptive speed update dynamically adjusts the particle speed to better guide the particle to move in the direction of the optimal solution. The chaotic position update calculation is performed based on the adaptive particle speed data, and the boundary constraints are performed by mirror mapping to ensure that the particles do not go beyond the boundary of the search space during the update process. Chaotic position update is a method to increase the randomness of the search by introducing chaotic variables, which can prevent particles from falling into the local optimal solution and improve the global search ability of the entire algorithm. Mirror mapping is to reflect the particle back to the inside of the search space when it crosses the boundary to ensure that the positions of all particles are within the valid range. After completing the position update, the position data of the target particle is subjected to multi-criteria convergence judgment to determine whether the current solution is close enough to the optimal solution. Through multi-criteria convergence judgment, multi-dimensional optimal solution set data is obtained, which represents a solution set that performs well on different targets. On this basis, a TOPSIS decision matrix is ​​constructed, and the control parameters of the heat exchanger are calculated by the entropy weight method. TOPSIS is a multi-criteria decision-making method that determines the optimal solution by comparing the distance between each solution and the ideal solution and the negative ideal solution. The entropy weight method is used to calculate the weight of each target, and can automatically adjust the weight according to the data distribution of each target to more reasonably balance the influence between the targets.The optimal control parameters of the heat exchanger are obtained by combining the TOPSIS method with the entropy weight method.

[0045] In an embodiment of the present invention, a dimensionless analysis model considering flow characteristics is established, combined with the distribution data of gas-liquid two-phase flow heat transfer coefficient, to accurately describe the flow and heat transfer characteristics inside the microchannel, thereby improving the accuracy of heat transfer performance prediction; a temperature field reconstruction method using multi-source data fusion is adopted, and temperature sensor data and phase change region heat transfer characteristic data are comprehensively utilized to achieve high-precision reconstruction of the heat exchanger temperature field; a deep learning prediction model with physical constraints is introduced, and the physical rationality of the temperature field prediction results is ensured through energy conservation constraints and attention mechanisms; a multi-objective optimization method based on an adaptive particle swarm algorithm is constructed, and the optimal control parameters that take into account both heat transfer efficiency and energy consumption are obtained through the entropy weight method and TOPSIS decision-making.

[0046] In a specific embodiment, the process of executing step S100 may specifically include the following steps:

[0047] Performing data acquisition and processing on the sensor array of the microchannel dual-phase precision phase change thermal module to obtain the original data of the temperature field, and performing spatial interpolation processing on the temperature sensing measurement signal according to the original data of the temperature field to obtain the temperature field acquisition data;

[0048] Sampling and processing the inlet and outlet pressure sensor data of the microchannel two-phase precision phase change heat exchange module to obtain pressure data, and sampling and processing the mass flow meter data of the microchannel two-phase precision phase change heat exchange module to obtain flow data;

[0049] According to the temperature measurement data of the thermocouple array in the phase change energy storage chamber inside the microchannel dual-phase precision phase change thermal module, the phase change state recognition calculation is performed to obtain the phase change completion data, and the phase change process is analyzed based on the phase change completion data to obtain the phase change start time data and phase change duration data;

[0050] The phase change start time data and the phase change duration data are substituted into the phase change energy storage calculation model to obtain the heat storage capacity data, and the characteristics of the phase change material are calculated based on the heat storage capacity data to obtain the phase change material characteristic data.

[0051] Specifically, the temperature data measured by the temperature sensor array inside the heat exchange module is collected. These sensors are arranged at different positions of the module to capture the details of the temperature field. The temperature data is spatially interpolated. Mathematical methods are used to estimate the temperature of areas not covered by sensors, making the entire temperature field data more complete and accurate. Methods such as Kriging interpolation or radial basis function interpolation are used. For example, if radial basis function interpolation is used, the temperature of a certain point in the temperature field is expressed as:

[0052] ;

[0053] in, is the temperature of the point where interpolation is required, is the number of sensors, is the interpolation coefficient, is the radial basis function, describing the measurement points With the point to be sought The functional relationship of the distance between them. Through this interpolation method, the blank areas in the temperature field are effectively filled to obtain complete temperature field acquisition data. The inlet and outlet pressure sensor data of the microchannel two-phase precision phase change heat exchange module are sampled and processed to obtain pressure data. The pressure sensor is arranged at the inlet and outlet of the heat exchange module to measure the pressure drop generated when the fluid flows through the microchannel. The pressure drop of the system reflects the resistance characteristics of the fluid in the channel, which affects the performance of the heat exchange process. Pressure drop Calculated by fluid dynamics formulas, such as:

[0054] ;

[0055] in, is the friction factor, is the length of the channel, is the hydraulic diameter of the microchannel, is the fluid density, is the flow velocity. Friction factor By empirical formula or by Reynolds number To determine, the Reynolds number is calculated by the formula:

[0056] ;

[0057] in is the dynamic viscosity of the fluid. Through these data and formulas, the flow state of the fluid in the microchannel is described. At the same time, the mass flow meter data of the microchannel two-phase precision phase change heat transfer module is sampled and processed to obtain flow data. The function of the mass flow meter is to measure the mass flow rate of the fluid. , directly affects the heat transfer coefficient and the efficiency of the phase change material. and fluid density The relationship represents the mass flow rate:

[0058] ;

[0059] This formula can convert volume flow into mass flow, which facilitates the calculation of energy exchange involved in the heat exchange process. In order to analyze the phase change process, a thermocouple array is arranged in the phase change energy storage chamber of the heat exchange module for temperature measurement. Based on the measured data, the phase change state identification calculation is performed to obtain the phase change completion data. The phase change completion indicates the degree to which the material completes the phase change during the phase change process. The different stages of the phase change process are determined by analyzing the temperature data measured by the thermocouple. Assumptions Indicates the phase transition completion, is the temperature of the phase change material, and the phase change completion is defined by normalizing the change of temperature over time:

[0060] ;

[0061] in and are the solid and liquid phase transition temperatures of the phase change material, respectively. The normalized formula helps the system determine the progress of the phase change, for example, when When the material is completely solid, and when When the material is completely transformed into liquid state, the phase change start time and phase change duration data are obtained by analyzing the phase change completion degree, which respectively represent the time point when the phase change starts and the duration of the entire phase change process. The phase change start time and phase change duration data are substituted into the phase change energy storage calculation model to obtain the heat storage capacity data. Phase change materials absorb or release a large amount of latent heat during the phase change process. Its heat storage capacity Calculated by the following formula:

[0062] ;

[0063] in is the mass of the phase change material, is the latent heat of phase change. This formula shows that the heat storage capacity of the phase change material depends on the mass of the material and the size of the latent heat of phase change. Through this step, the performance of the phase change material in the heat storage process is quantified. Based on the heat storage capacity data, the characteristics of the phase change material are calculated to obtain the characteristic data of the phase change material. The characteristic data of the phase change material includes its thermal conductivity Specific heat capacity and parameters such as latent heat of phase change, which help describe the thermal properties of materials during phase change. For example, thermal conductivity Describes the ability of a material to transfer heat, expressed by Fourier's law of heat conduction:

[0064] ;

[0065] in is the heat flux density, is the temperature gradient. This formula shows that the greater the thermal conductivity of a material, the faster its heat transfer rate. Describe the heat required by the material during the heating process, calculated by the following formula:

[0066] ;

[0067] in For the absorbed heat, is the change in temperature. By calculating these physical parameters, the thermodynamic behavior of phase change materials at different stages can be reflected.

[0068] In a specific embodiment, the process of executing step S200 may specifically include the following steps:

[0069] The Reynolds number, Weber number and Kapitza number are calculated based on the pressure data and flow data to obtain dimensionless data of flow characteristics in the microchannel;

[0070] The dimensionless data of flow characteristics in the microchannel are judged by intervals. When the Weber number is less than 0.5, it is judged as an intermittent slug flow area. When the Weber number is between 0.5 and 2.5, it is judged as an annular flow area. When the Weber number is greater than 2.5, it is judged as a mist flow area. The flow pattern area judgment data is obtained.

[0071] Calculating the flow pattern conversion characteristic parameters according to the flow pattern area determination data to obtain the flow pattern criterion value, and calculating the boundary layer thickness of the flow pattern criterion value to obtain the boundary layer characteristic data in the microchannel;

[0072] According to the boundary layer characteristic data and flow pattern criterion values ​​in the microchannel, the bubble slip coefficient is calculated for the intermittent elastic flow area to obtain the bubble slip characteristic data, and the annular flow liquid film fluctuation is calculated for the flow pattern criterion values ​​to obtain the liquid film fluctuation characteristic data;

[0073] The droplet entrainment coefficient in the mist flow area is calculated according to the flow type criterion value to obtain the droplet entrainment characteristic data. The bubble slip characteristic data, liquid film fluctuation characteristic data and droplet entrainment characteristic data are substituted into the two-phase flow heat transfer coefficient calculation model to obtain the gas-liquid two-phase flow heat transfer coefficient distribution data.

[0074] Specifically, the Reynolds number, Weber number and Kapitza number are calculated based on the pressure data and flow data to obtain dimensionless data of the flow characteristics in the microchannel. , which represents the proportional relationship between the inertial force and the viscous force of the fluid in the pipeline, and is an important indicator for determining whether the flow state is laminar or turbulent. The calculation formula of Reynolds number is:

[0075] ;

[0076] in, is the density of the fluid, is the flow rate, is the hydraulic diameter of the microchannel, is the dynamic viscosity of the fluid. The Reynolds number is a basic parameter used to determine the flow type: When , the flow is usually considered to be laminar, and when When the flow enters the turbulent state, the Reynolds number is generally low in microchannels, and the flow is mainly in the laminar or transitional region. The Weber number is calculated based on factors such as the inertial force and surface tension of the fluid. , which is used to measure the relationship between the inertial force and surface tension of a fluid. The Weber number is calculated as:

[0077] ;

[0078] in, is the fluid density, is the flow rate, is the hydraulic diameter of the microchannel, is the surface tension coefficient of the fluid. The Weber number reflects the stability of droplets or bubbles in the fluid flow. When the Weber number is small, the surface tension plays a dominant role, and the bubbles and droplets in the flow are more stable. A larger Weber number means that the inertial force is dominant, and the bubbles or droplets are easier to break or disperse. At the same time, the Kapitsa number is calculated. The Kapitza number is used to describe the interaction between gravity, surface tension and viscosity, and is particularly important in the phase change heat transfer process. The calculation formula of the Kapitza number is:

[0079] ;

[0080] in, is the acceleration due to gravity, is the fluid density, is the hydraulic diameter, is the dynamic viscosity. The Kapitza number is particularly important in describing the stability of the liquid film during the phase change process. The larger its value, the more stable the liquid film is, which helps to improve the heat transfer efficiency. According to the calculated Reynolds number, Weber number and Kapitza number, the dimensionless data of the flow characteristics in the microchannel are obtained, and the flow morphology is determined according to the dimensionless data. When the Weber number is less than 0.5, the flow is judged to be an intermittent slug flow region. In this flow state, the liquid and gas alternately form a fluid structure similar to a projectile, with strong interface fluctuations, which is suitable for phase change heat transfer. When the Weber number is between 0.5 and 2.5, it is judged to be an annular flow region. In this state, the liquid forms an annular liquid film along the tube wall, and the gas flows through the center part. The volatility of the liquid film has a significant effect on the heat transfer process. If the Weber number is greater than 2.5, the flow is judged to be a mist flow region. At this time, the liquid is dispersed in the airflow in the form of small droplets. This flow morphology is conducive to the formation of efficient gas-liquid two-phase heat exchange. The flow pattern conversion characteristic parameters are calculated based on the flow pattern area determination data to obtain the flow pattern criterion value. The flow pattern criterion value can describe the conversion characteristics between different flow patterns, such as the transition conditions from intermittent slug flow to annular flow. These values ​​are obtained through empirical formulas or experimental data regression. Based on the flow pattern criterion value, the boundary layer thickness in the microchannel is calculated. Boundary layer thickness It indicates the area where the fluid in the flow is significantly affected by the wall. The calculation formula is usually:

[0081] ;

[0082] in is the hydraulic diameter, is the Reynolds number. The calculation of boundary layer thickness helps to understand the heat transfer characteristics between the fluid and the wall in the microchannel, because the thickness of the boundary layer determines the magnitude of heat transfer and flow resistance. According to the boundary layer characteristic data and flow type criterion value in the microchannel, the bubble slip coefficient is calculated for the intermittent elastic flow area to obtain the bubble slip characteristic data. Bubble slip coefficient ( ) is used to describe the relative movement speed of the gas phase in the liquid phase, and its calculation formula is:

[0083] ;

[0084] in, and are the flow rates of the gas phase and the liquid phase, respectively, is the average flow velocity of the mixed phase. The bubble slip phenomenon has a significant effect on the heat transfer performance of intermittent slug flow, because the slip can increase the relative velocity between gas and liquid, thereby enhancing the heat transfer effect. At the same time, based on the flow type criterion value, the liquid film fluctuation characteristics of the annular flow area are calculated to obtain the liquid film fluctuation characteristic data. Liquid film fluctuation has an important influence on the heat transfer of annular flow, because the fluctuation can increase the surface area of ​​the liquid film, thereby enhancing the heat transfer capacity. The wavelength of the liquid film fluctuation ( ) is estimated by the following formula:

[0085] ;

[0086] in, is the surface tension, is the fluid density, is the flow rate. A smaller wavelength indicates a higher wave frequency of the liquid film, which increases the heat transfer area and improves the heat transfer efficiency. For the mist flow area, the droplet entrainment coefficient is calculated according to the flow pattern criterion value to obtain the droplet entrainment characteristic data. Droplet entrainment coefficient Indicates the degree to which droplets are carried by the airflow, and its calculation formula is:

[0087] ;

[0088] in, is the air flow resistance of the droplet, is the gas phase density, is the gas phase flow rate, is the windward area of ​​the droplets. Droplet entrainment has an important influence on the heat transfer of mist flow, because a large number of droplets dispersed in the airflow can significantly increase the contact area between the gas and liquid phases and thus improve the heat transfer efficiency. Substituting the bubble slip characteristic data, liquid film fluctuation characteristic data and droplet entrainment characteristic data into the two-phase flow heat transfer coefficient calculation model, the distribution data of the gas-liquid two-phase flow heat transfer coefficient is obtained. Heat transfer coefficient The calculation of is realized by the following relationship:

[0089] ;

[0090] in , , is a constant related to the flow pattern, is the Prandtl number, which is defined as:

[0091] ;

[0092] in, is the specific heat capacity, is the dynamic viscosity, is the thermal conductivity. The Prandtl number describes the relative rates of momentum diffusion and heat diffusion in a fluid.

[0093] In a specific embodiment, the process of executing step S300 may specifically include the following steps:

[0094] The heat storage capacity data in the phase change material characteristic data is processed piecewise linearly to obtain piecewise linear phase change characteristic data, and the phase change region is meshed according to the gas-liquid two-phase flow heat transfer coefficient distribution data to obtain the phase change region mesh data;

[0095] Substitute the piecewise linearized phase change characteristic data into the heat source heat removal cycle and refrigeration cycle coupling model to obtain the cycle coupling heat transfer equation, and perform natural convection term correction calculation on the cycle coupling heat transfer equation, and establish the phase change heat transfer control equation, which is:

[0096] ρCp(∂T / ∂t)+ρCpu·∇T=∇·(k∇T)+ρL(∂f / ∂t)+βgρCp(T-T0);

[0097] Among them, ρ is the density of phase change material; Cp is the specific heat capacity of phase change material; T is temperature field; t is time; u is velocity field; k is thermal conductivity; L is latent heat of phase change; f is liquid fraction; β is volume thermal expansion coefficient; g is gravitational acceleration; T0 is reference temperature;

[0098] The phase change heat transfer control equation is discretized according to the grid data of the phase change region to obtain a discretized control equation group, and the discretized control equation group is solved by time advancement to obtain the temperature field data of the phase change region;

[0099] The heat flux density distribution in the phase change region is calculated according to the temperature field data of the phase change region to obtain the heat flux density distribution data, and the temperature field data and the heat flux density distribution data of the phase change region are comprehensively processed to obtain the heat transfer characteristic data of the phase change region.

[0100] Specifically, the heat storage capacity data in the phase change material characteristic data is processed by piecewise linearization, and the complex nonlinear heat storage characteristics are divided into several linear intervals, and the characteristics in each interval are considered to be linear. The piecewise linearization process allows the phase change material to be described by a simple linear relationship in each small temperature interval, simplifies the complexity of numerical calculations, and ensures the accuracy of calculations to a certain extent. For example, suppose there is a phase change material whose heat storage capacity It is a nonlinear function, which is linearized in different temperature ranges, such as arrive Use linear function in the interval To indicate that in the interval arrive Internal use To represent. According to the distribution data of the heat transfer coefficient of the gas-liquid two-phase flow, the phase change region is meshed to obtain the mesh data of the phase change region. The phase change region is spatially discretized and the continuous heat transfer process is divided into multiple small discrete units to facilitate the calculation of the heat transfer and flow characteristics of each small area. In numerical simulation, the finite difference method, finite volume method or finite element method is used to mesh the region. For example, the entire phase change region is divided into Small units, the volume of each small unit is , and establish a local heat transfer balance equation for each unit. Substitute the piecewise linearized phase change characteristic data into the heat source heat removal cycle and refrigeration cycle coupling model to obtain the cyclic coupled heat transfer equation. This coupling model comprehensively considers the two stages of heat source (heating) and refrigeration (cooling) of the phase change material in the heat exchange process. Through the coupling model, the heat transfer behavior of the phase change material in these two stages is described at the same time to ensure a complete description of the entire heat exchange process. In the process of establishing the cyclic coupled heat transfer equation, the natural convection term is corrected. Natural convection has a significant effect on heat transfer, and its effect usually depends on the temperature difference and density difference of the fluid. In the heat transfer equation, the buoyancy term is introduced to describe the influence of natural convection. The corrected phase change heat transfer control equation is:

[0101] ;

[0102] in, is the density of the phase change material, is the specific heat capacity of the phase change material, is the temperature field, For time, is the velocity field, is the thermal conductivity, is the latent heat of phase change, is the liquid fraction, is the volume thermal expansion coefficient, is the acceleration due to gravity, is the reference temperature. Each term in the equation has physical meaning. The first term represents the change of temperature field with time, the second term is the convective heat transfer term, the third term represents the heat conduction process, the fourth term describes the release or absorption of latent heat of phase change, and the fifth term is the buoyancy term of natural convection, which describes the buoyancy effect caused by temperature difference. The phase change heat transfer control equation is discretized for numerical solution. The continuous partial differential equation is converted into a set of algebraic equations in discrete form so that it can be solved by numerical methods. The finite difference method is used for discretization. For example, the time derivative part is It can be expressed by forward difference as:

[0103] ;

[0104] in and Respectively expressed in time and The temperature of the moment, is the time step. Spatial derivative part Discretization is performed through central difference to obtain a discretized set of control equations. These discretized sets of equations are solved by iterative methods, such as using explicit or implicit time-marching methods to gradually solve the temperature field for each time step until the temperature field distribution of the entire phase change region is obtained. The heat flux density distribution in the phase change region is calculated based on the temperature field data of the phase change region. Heat flux density It is an important parameter to measure the heat transfer rate, which indicates the heat flow per unit area. Its calculation formula is:

[0105] ;

[0106] in is the thermal conductivity, is the temperature gradient. The temperature gradient in each grid unit is calculated through the temperature field data to obtain the corresponding heat flux density distribution data. The distribution of heat flux density can intuitively reflect how heat is transferred from the high temperature area to the low temperature area during the phase change process, thereby helping to understand the thermodynamic behavior of phase change materials. The temperature field data and heat flux density distribution data of the phase change region are comprehensively processed to obtain the heat transfer characteristic data of the phase change region. For example, by calculating the temperature change rate and the change of heat flux density at different time and space positions, the transient heat transfer coefficient distribution of the phase change region is obtained.

[0107] In a specific embodiment, the process of executing step S400 may specifically include the following steps:

[0108] According to the temperature field acquisition data and the radial basis function interpolation algorithm, a radial basis function matrix and a distance weight matrix are constructed to obtain an initial interpolation matrix, and the heat transfer characteristic data of the phase change region are normalized to obtain normalized heat transfer characteristic data;

[0109] Substitute the normalized heat transfer characteristic data into the initial interpolation matrix to obtain the corrected interpolation coefficient matrix, and construct the temperature field reconstruction objective function based on the corrected interpolation coefficient matrix. The temperature field reconstruction objective function is:

[0110] F(G)=||ΦG-Y||²+λ||WG||²;

[0111] Among them, F(G) is the temperature field reconstruction objective function, Φ is the radial basis function interpolation matrix; G is the temperature field to be reconstructed; Y is the measured temperature data; λ is the regularization parameter; W is the Laplace operator;

[0112] Performing gradient calculation on the temperature field reconstruction objective function data to obtain objective function gradient data, and performing conjugate direction update calculation based on the objective function gradient data to obtain search direction data;

[0113] Substitute the search direction data into the linear search algorithm to obtain the optimal step length data, and iteratively update the temperature field based on the optimal step length data. When the iterative residual is less than 10⁻ 6 The iteration is stopped when , and the temperature field reconstruction data is obtained.

[0114] Specifically, the radial basis function matrix and the distance weight matrix are constructed according to the temperature field acquisition data and the radial basis function interpolation algorithm to obtain the initial interpolation matrix. The radial basis function is an interpolation tool suitable for the interpolation and reconstruction of multidimensional data. Assume that Known measurement points , the temperature field acquisition data is , radial basis function matrix Each element of is represented by a radial basis function, for example, the Gaussian radial basis function is:

[0115] ;

[0116] in, Indicates the point to be interpolated and measuring points The Euclidean distance between is a parameter that controls the range of the function. By substituting the distances between all measured points into the radial basis function, a The matrix , which is used to describe the spatial relationship between different sampling points. While constructing the radial basis function matrix, the distance weight matrix is ​​constructed , which is used to quantify the contribution of different sampling points to the interpolation points. Each element of the weight matrix Calculated by the reciprocal of the distance, for example:

[0117] ;

[0118] in, is a small constant used to prevent division by zero errors. The distance weight matrix obtained helps to assign lower weights to sampling points far from the interpolation points, thereby improving the accuracy of interpolation. The heat transfer characteristic data of the phase change region are normalized to obtain normalized heat transfer characteristic data. Substituting the normalized heat transfer characteristic data into the initial interpolation matrix, the modified interpolation coefficient matrix is ​​obtained. The corrected interpolation coefficient matrix is ​​used to establish the reconstruction model of the temperature field and construct the temperature field reconstruction objective function:

[0119] ;

[0120] in, represents the objective function of temperature field reconstruction, is the radial basis function interpolation matrix, is the temperature field to be reconstructed, is the measured temperature data, is the regularization parameter, is the Laplace operator. The objective function consists of two parts: the first The difference between the reconstructed temperature field and the measured temperature data represents the accuracy of the interpolation; the second term The regularization term is used to control the smoothness of the temperature field by introducing the Laplace operator to avoid overfitting during the reconstruction process. Controls the trade-off between interpolation accuracy and smoothness. The value makes the temperature field smoother, while smaller The value makes the reconstruction result closer to the original data. The gradient of the objective function is calculated to obtain the gradient data of the objective function. The gradient of the objective function is calculated by about The partial derivative calculation of is:

[0121] ;

[0122] in, is the transpose of the radial basis function interpolation matrix, is the transpose of the Laplace operator. The gradient data indicates the direction of change of the objective function in the parameter space. By calculating the gradient, we can determine how to adjust the temperature field. To minimize the objective function. Based on the gradient data of the objective function, the conjugate direction update calculation is performed to obtain the search direction data. The conjugate gradient method is an effective iterative method for solving large linear systems. By selecting a search direction that is conjugate to the previous direction in each iteration, the optimal solution of the objective function can be quickly found. The update formula for the search direction is:

[0123] ;

[0124] in, For the The search direction of the iteration, For the The gradient at iteration , is the conjugate coefficient, calculated using the Fletcher-Reeves formula:

[0125] ;

[0126] The role of the conjugate coefficient is to ensure that the new search direction is orthogonal to the previous search direction in a certain sense, thereby accelerating convergence. Substitute the search direction data into the linear search algorithm to find the optimal step length data. The goal of the linear search is to find an optimal step length in the current search direction so that the value of the objective function is minimized after moving in this direction. The calculation of the step length is expressed by the following formula:

[0127] ;

[0128] in, For the The optimal step length of the iteration. Through linear search, it is ensured that each step update is carried out along the direction where the objective function drops fastest, thereby accelerating the convergence speed. Based on the optimal step length data, the temperature field Perform iterative updates:

[0129] ;

[0130] By gradually updating the temperature field, the minimum value of the objective function is gradually approached. After each iteration, the residual is calculated , when the residual is less than the specified threshold When , the iteration is stopped, which means that the reconstruction result of the temperature field is accurate enough and meets the predetermined accuracy requirement, and the final temperature field reconstruction data is obtained.

[0131] In a specific embodiment, the process of executing step S500 may specifically include the following steps:

[0132] The temperature field reconstruction data is input into the encoding layer of the temperature field evolution prediction model for processing to obtain the temperature field feature encoding data. The encoding layer includes three convolution sublayers. The convolution kernel size of the first convolution sublayer is 3×3, the step size is 1, and the number of output channels is 64. The convolution kernel size of the second convolution sublayer is 3×3, the step size is 2, and the number of output channels is 128. The convolution kernel size of the third convolution sublayer is 3×3, the step size is 2, and the number of output channels is 256.

[0133] The temperature field feature encoding data is input into the long short-term memory network layer of the temperature field evolution prediction model to extract time series features. The long short-term memory network layer includes four parallel memory units, each of which includes an input gate, a forget gate, and an output gate. The number of neurons in the hidden layer is 512, and the time series feature data is obtained;

[0134] Energy conservation constraint calculation is performed on the time series feature data to obtain the constrained time series feature data, and the constrained time series feature data is input into the attention mechanism layer of the temperature field evolution prediction model for spatial correlation analysis to obtain spatial attention feature data. The attention mechanism layer adopts a self-attention structure, the number of attention heads is 8, and the hidden layer dimension is 256;

[0135] The spatial attention feature data is input into the decoding layer of the temperature field evolution prediction model to reconstruct the temperature field and obtain the temperature field prediction data. The decoding layer includes three deconvolution sublayers. The convolution kernel size of the first deconvolution sublayer is 4×4, the step size is 2, and the number of output channels is 128. The convolution kernel size of the second deconvolution sublayer is 4×4, the step size is 2, and the number of output channels is 64. The convolution kernel size of the third deconvolution sublayer is 3×3, the step size is 1, and the number of output channels is 1.

[0136] Specifically, the temperature field data is effectively encoded. The encoding layer consists of three convolutional sublayers. The function of each convolutional sublayer is to gradually extract high-level features in the data for subsequent time series analysis and spatial correlation modeling. The convolution kernel size of the first convolutional sublayer is , the step size is 1, and the number of output channels is 64. Its main function is to extract fine-grained features from the input temperature field data. Suppose the input temperature field reconstruction data is ,in , represent the height and width of the temperature field respectively, is the number of channels. The convolution operation is expressed by the following formula:

[0137] ;

[0138] in, is the convolution kernel, is the bias, is the output feature value after convolution. Through this convolution operation, local spatial information is extracted and the spatial resolution of the input is maintained. The preliminary features extracted in the first convolution sublayer enter the second convolution sublayer, and the convolution kernel size of this sublayer is also , the step size is set to 2, and the number of output channels is increased to 128. The increase in step size reduces the size of the feature map by half, thereby gradually aggregating a larger receptive field and extracting more global information. Through convolution processing, the spatial structure of the features is retained, the redundancy of the data is reduced, and higher-level abstract features are extracted. Then enter the third convolution sublayer, and its convolution kernel size is still , the step size is 2, and the number of output channels is 256. This convolution sublayer increases the number of channels to increase the feature expression capability, while reducing the size of the feature map, so that the model focuses on more abstract features. The design of increasing the number of channels layer by layer helps to improve the model's ability to represent temperature field features, making subsequent time series analysis more representative and robust. The temperature field feature encoding data is input into the long short-term memory network (LSTM) layer of the temperature field evolution prediction model for time series feature extraction. The LSTM network is used to process time series data and capture the change pattern of the temperature field over time. The LSTM layer consists of four parallel memory cells, each of which contains an input gate, a forget gate, and an output gate. These gate structures work together to control the flow of information. The input gate is used to select which new information needs to be stored in the memory cell, the forget gate is used to select the old information that needs to be forgotten, and the output gate determines which information needs to be passed to the next moment. The state update formula of LSTM is:

[0139] ;

[0140] ;

[0141] ;

[0142] ;

[0143] ;

[0144] in, , , are the activation values ​​of the forget gate, input gate, and output gate, respectively. is the sigmoid activation function, is the hyperbolic tangent function, and are weight matrices and bias terms, respectively, and © represents element-wise multiplication. Through these gate structures, LSTM can effectively capture the temporal dependency in the temperature field data, and the hidden state of the output As time series feature data. Energy conservation constraint calculation is performed on the time series feature data. Energy conservation is a physical constraint that needs to be followed during the evolution of the temperature field to ensure that the predicted temperature field conforms to the laws of thermodynamics. The energy conservation constraint is achieved by balancing the temperature change at each moment. The specific formula is:

[0145] ;

[0146] in, For the The quality of the unit, is the specific heat capacity, is the rate of change of temperature with time, and are the input and output heat respectively. By performing energy conservation constraints on the time series feature data, the constrained time series feature data is obtained. The constrained time series feature data is input into the attention mechanism layer of the temperature field evolution prediction model for spatial correlation analysis. The attention mechanism layer adopts a self-attention structure, which calculates the correlation between different positions in the temperature field to enhance the model's attention to key areas. The calculation of the self-attention mechanism includes the following steps: calculating the query, key, and value matrices, and obtaining the attention weight through the following formula:

[0147] ;

[0148] in, , , are query, key, and value matrices respectively, is the dimension of the key, which is used to scale the attention score, and the softmax function is used to normalize the weights so that the sum of the weights is 1. With 8 attention heads, the model can learn the spatial relationship of the temperature field from 8 different subspaces and enhance the modeling ability of complex correlations. The dimension of the hidden layer is set to 256 to ensure that there is enough capacity to capture the global characteristics of the temperature field. The spatial attention feature data is input into the decoding layer of the temperature field evolution prediction model for temperature field reconstruction. The decoding layer consists of three deconvolution sublayers, which are used to restore the extracted features to the original temperature field size and structure. The convolution kernel size of the first deconvolution sublayer is 4×4, the stride is 2, and the number of output channels is 128. The resolution of the feature map is increased by upsampling. The convolution kernel size of the second deconvolution sublayer is also 4×4, the stride is 2, and the number of output channels is reduced to 64. It further upsamples and reduces the number of feature channels, so that the feature map gradually approaches the form of the original input. The convolution kernel size of the third deconvolution sublayer is 3×3, the step size is 1, and the number of output channels is 1, which is used to finally restore the feature map to a single-channel temperature field image. After processing by the decoding layer, the predicted data of the temperature field are finally obtained, which is used to predict the evolution of the temperature field in the future time period.

[0149] In a specific embodiment, the process of executing step S600 may specifically include the following steps:

[0150] According to the temperature field prediction data, the first optimization objective function H1(x)=(Nin-Nout) / (Nin-N∞) and the second optimization objective function H2(x)=eQΔP are constructed respectively, where: H1(x) represents the first optimization objective function, Nin is the inlet temperature; Nout is the outlet temperature; H2(x) represents the second optimization objective function, N∞ is the ambient reference temperature; e is the fluid density; Q is the volume flow rate; ΔP is the system pressure drop;

[0151] The dual-objective optimization function H(x)=w1·H1(x)+w2·H2(x) is obtained by weighted summation method, where w1 and w2 are weight coefficients of heat transfer efficiency and energy consumption, respectively, and w1+w2=1;

[0152] Based on the dual-objective optimization function, the first type of particles are initialized and the second type of particles are initialized to obtain the initial particle swarm data of the double-layer structure;

[0153] The initial particle swarm data of the double-layer structure is input into the hierarchical fitness calculation model to obtain two layers of fitness. The first layer of the hierarchical fitness calculation model calculates the heat transfer efficiency; the second layer calculates the energy consumption;

[0154] The fitness of the two layers is standardized by the minimum-maximum normalization method to obtain the comprehensive fitness data, and non-dominated sorting is performed according to the comprehensive fitness data to obtain the hierarchical optimal position data;

[0155] Perform adaptive speed update calculation on the hierarchical optimal position data to obtain adaptive particle speed data, and perform chaotic position update calculation based on the adaptive particle speed data, perform boundary constraints through mirror mapping, and obtain target particle position data;

[0156] Multi-criteria convergence judgment is performed on the target particle position data to obtain multi-dimensional optimal solution set data. The TOPSIS decision matrix is ​​constructed based on the multi-dimensional optimal solution set data, and the heat exchanger control parameters are calculated by the entropy weight method.

[0157] Specifically, two optimization objective functions are constructed according to the temperature field prediction data, namely the first optimization objective function and the second optimization objective function. The form is:

[0158] ;

[0159] in, is the inlet temperature, is the outlet temperature, is the ambient reference temperature. The physical meaning of this function is to measure the efficiency of the heat exchange process. represents the degree of temperature reduction of the fluid in the heat exchanger, and This is the theoretically maximum possible temperature drop. It indicates the ratio of the heat transfer efficiency in the actual heat transfer process to the theoretical optimum. The second optimization objective function It is used to measure energy consumption, specifically:

[0160] ;

[0161] in, is the density of the fluid, is the volume flow rate, is the system pressure drop. This function describes the energy consumption of the heat exchange system per unit time. It reflects the energy required for the fluid to overcome resistance in the channel, and It represents the energy consumed in the entire process of fluid flowing through the heat exchanger. Represents the energy consumption characteristics of the system. The goal is to minimize energy consumption while ensuring heat transfer efficiency. The two objective functions are combined by weighted summation to obtain the dual-objective optimization function :

[0162] ;

[0163] in, and are the weight coefficients of heat transfer efficiency and energy consumption, respectively, and satisfy By adjusting and The value of , trade-off between heat transfer efficiency and energy consumption, for example, if you pay more attention to heat transfer efficiency, increase The weight of The weight of . Particle swarm optimization is performed based on a dual-objective optimization function. The first type of particles and the second type of particles are initialized to obtain the initial particle swarm data of the double-layer structure. Particle swarm optimization is a global optimization algorithm based on swarm intelligence, and each particle represents a possible solution. The first type of particles are used to represent parameters related to heat transfer efficiency, and the second type of particles are used to represent parameters related to energy consumption. During initialization, the initial position and velocity of the particles are usually generated by random numbers to ensure the coverage of the search space. The initial particle swarm data of the double-layer structure is input into the hierarchical fitness calculation model to evaluate the performance of each particle. The first layer of the fitness calculation model is used to calculate the heat transfer efficiency, that is, according to The value of measures the performance of each particle in improving heat transfer efficiency. The higher the fitness, the better the parameter combination corresponding to the particle performs in terms of heat transfer efficiency. The second layer is used to calculate energy consumption, that is, according to The value of is used to evaluate energy consumption. The lower the fitness, the lower the energy consumption. Through the hierarchical fitness calculation, the performance of each particle in terms of heat transfer efficiency and energy consumption is evaluated at the same time. In order to standardize the two fitness values ​​so that they can be compared uniformly, the minimum-maximum normalization method is used to standardize the two-layer fitness data. The normalization formula is:

[0164] ;

[0165] in, For the The fitness value of a particle, and are the minimum and maximum fitness values ​​of all particles, respectively. The normalized fitness value is in the range of [0,1], which effectively eliminates the influence of dimension and enables fair comparison between different objectives. Non-dominated sorting is performed based on the standardized comprehensive fitness data to obtain hierarchical optimal position data. Non-dominated sorting is a sorting method for multi-objective optimization. Particles are stratified according to their performance on different objectives. Particles that perform well on all objectives are divided into the first layer, the second best particles are divided into the second layer, and so on. In this way, it is possible to identify which particles have better comprehensive performance on multiple objectives, thereby providing a reference for the optimization direction. An adaptive speed update calculation is performed on the hierarchical optimal position data to obtain adaptive particle speed data. The speed update formula in particle swarm optimization is:

[0166] ;

[0167] in, For the The particle in The speed of generation, is the inertia weight, and is the learning factor, and is a random number, is the historical optimal position of the particle, is the global optimal position. Through the adaptive update of velocity, the particles approach the optimal solution at a suitable pace in the solution space, thereby improving the optimization efficiency. After completing the velocity update, the chaotic position update calculation is performed based on the adaptive particle velocity data, and the boundary constraints are performed through mirror mapping. Chaotic position update is a method of increasing the randomness of the search by introducing chaotic variables to prevent the particle swarm from falling into the local optimum. Mirror mapping is used to reflect the particle position back to the solution space when it crosses the boundary, ensuring that the updates of all particles are within the valid range. For example, if the position of a particle exceeds the upper limit , then adjust its position to . Perform multi-criteria convergence judgment on the target particle position data to obtain multi-dimensional optimal solution set data. The purpose of multi-criteria convergence judgment is to determine whether the termination condition of optimization is met, for example, whether the speed of all particles is small enough, or whether the fitness value is stable and no longer changes. When the conditions are met, the current particle position is used as the optimal solution set. Construct a TOPSIS decision matrix based on the multi-dimensional optimal solution set data to select the final control parameters. TOPSIS is a multi-criteria decision-making method that evaluates the quality of each solution by calculating the distance between each solution and the ideal solution and the negative ideal solution. Calculate the Euclidean distance of each solution from the ideal solution (optimal value of each indicator) and the negative ideal solution (worst value of each indicator):

[0168] ;

[0169] in, For the The solution indicators, and are the ideal solution and the negative ideal solution respectively. Then calculate the relative proximity:

[0170] ;

[0171] The greater the relative closeness, the closer the solution is to the ideal solution. In this way, the optimal heat exchanger control parameters are selected. The weight of each indicator is calculated by the entropy weight method to ensure that the influence of different indicators can be reasonably reflected. The entropy weight method is a weight calculation method based on information entropy, which automatically adjusts the weight according to the discrete degree of the data, thereby avoiding the subjectivity of artificially setting the weight.

[0172] See also Figure 2 , Figure 2 A schematic block diagram of the structure of the temperature field analysis device 200 of the heat exchange module provided in the embodiment of the present application is shown in FIG. Figure 2 As shown, the temperature field analysis device 200 of the heat exchange module includes:

[0173] The acquisition module 210 is used to perform multi-source data acquisition on the microchannel dual-phase precision phase change heat transfer module to obtain temperature field acquisition data, pressure data, flow data and phase change material characteristic data;

[0174] The analysis module 220 is used to perform dimensionless number analysis and calculation on the pressure data and the flow data to obtain the flow pattern criterion value, and calculate the gas-liquid two-phase flow heat transfer coefficient distribution data according to the flow pattern criterion value;

[0175] The calculation module 230 is used to substitute the gas-liquid two-phase flow heat transfer coefficient distribution data and the phase change material characteristic data into the phase change heat transfer control equation for coupling calculation to obtain the phase change region heat transfer characteristic data;

[0176] Iteration module 240, used to input radial basis function interpolation algorithm to perform conjugate gradient iterative calculation according to heat transfer characteristic data of phase change region and temperature field acquisition data, so as to obtain temperature field reconstruction data;

[0177] Prediction module 250, used for inputting temperature field reconstruction data into temperature field evolution prediction model to perform temperature field evolution prediction and obtain temperature field prediction data;

[0178] The solution module 260 is used to establish a dual-objective optimization function including heat exchange efficiency and energy consumption according to the temperature field prediction data, and obtain the heat exchanger control parameters through iterative solution by particle swarm algorithm.

[0179] Through the collaborative cooperation of the above-mentioned components, by establishing a dimensionless analysis model that takes flow pattern characteristics into consideration and combining the distribution data of the heat transfer coefficient of gas-liquid two-phase flow, the flow and heat transfer characteristics inside the microchannel are accurately described, thereby improving the accuracy of the heat transfer performance prediction; a temperature field reconstruction method based on multi-source data fusion is adopted, which comprehensively utilizes temperature sensor data and phase change area heat transfer characteristic data to achieve high-precision reconstruction of the heat exchanger temperature field; a deep learning prediction model with physical constraints is introduced, and the physical rationality of the temperature field prediction results is ensured through energy conservation constraints and attention mechanisms; a multi-objective optimization method based on an adaptive particle swarm algorithm is constructed, and the optimal control parameters that take into account both heat transfer efficiency and energy consumption are obtained through the entropy weight method and TOPSIS decision-making.

[0180] See also Figure 3 , Figure 3 The present invention provides a schematic block diagram of the structure of a temperature field analysis device 300 for a heat exchange module according to an embodiment of the present application. The temperature field analysis device 300 for a heat exchange module includes a processor 301 and a memory 302. The processor 301 and the memory 302 are connected via a device bus 303, wherein the memory 302 may include a non-volatile storage medium and an internal memory.

[0181] The non-volatile storage medium can store a computer program. The computer program includes program instructions, and when the program instructions are executed by the processor 301, the processor 301 can execute any of the above-mentioned temperature field analysis methods of the heat exchange module.

[0182] The processor 301 is used to provide computing and control capabilities to support the operation of the temperature field analysis device 300 of the entire heat exchange module.

[0183] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor 301, the processor 301 can execute any of the above-mentioned temperature field analysis methods of the heat exchange module.

[0184] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the scheme of the present application, and does not constitute a limitation on the temperature field analysis device 300 of the heat exchange module involved in the scheme of the present application. The temperature field analysis device 300 of the specific heat exchange module may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0185] It should be understood that the processor 301 may be a central processing unit (CPU), and the processor 301 may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0186] It should be noted that technicians in the relevant field can clearly understand that for the convenience and simplicity of description, the specific working process of the temperature field analysis equipment 300 of the heat exchange module described above can refer to the corresponding process of the temperature field analysis method of the aforementioned heat exchange module, and will not be repeated here.

[0187] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by one or more processors, the one or more processors implement the temperature field analysis method of the heat exchange module provided in the embodiment of the present application.

[0188] The computer-readable storage medium may be an internal storage unit of the temperature field analysis device 300 of the heat exchange module in the aforementioned embodiment, such as a hard disk or memory of the temperature field analysis device 300 of the heat exchange module. The computer-readable storage medium may also be an external storage device of the temperature field analysis device 300 of the heat exchange module, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc., equipped with the temperature field analysis device 300 of the heat exchange module.

[0189] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0190] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0191] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A temperature field analysis method for a heat exchange module, characterized in that: include: Perform multi-source data collection on the microchannel dual-phase precision phase change heat transfer module to obtain temperature field collection data, pressure data, flow data and phase change material characteristic data; Performing dimensionless number analysis and calculation on the pressure data and the flow data to obtain flow pattern criterion values, and calculating gas-liquid two-phase flow heat transfer coefficient distribution data according to the flow pattern criterion values; Substituting the gas-liquid two-phase flow heat transfer coefficient distribution data and the phase change material characteristic data into the phase change heat transfer control equation for coupling calculation to obtain the heat transfer characteristic data of the phase change region; specifically including: performing piecewise linearization processing on the heat storage capacity data in the phase change material characteristic data to obtain piecewise linearized phase change characteristic data, and meshing the phase change region according to the gas-liquid two-phase flow heat transfer coefficient distribution data to obtain the phase change region mesh data; substituting the piecewise linearized phase change characteristic data into the heat source heat removal cycle and refrigeration cycle coupling model to obtain the cycle coupling heat transfer equation, and performing natural convection term correction calculation on the cycle coupling heat transfer equation, and establishing the phase change heat transfer control equation, the phase change heat transfer control equation is: ;in, is the density of phase change material; is the specific heat capacity of the phase change material; is the temperature field; For time; is the velocity field; is the thermal conductivity; is the latent heat of phase change; is the liquid fraction; is the volume thermal expansion coefficient; is the acceleration due to gravity; is a reference temperature; discretizing the phase change heat transfer control equation according to the phase change region grid data to obtain a discretized control equation group, and performing time-advancing solution on the discretized control equation group to obtain the temperature field data of the phase change region; calculating the heat flux density distribution in the phase change region according to the phase change region temperature field data to obtain the heat flux density distribution data, and comprehensively processing the phase change region temperature field data and the heat flux density distribution data to obtain the heat transfer characteristic data of the phase change region; According to the heat transfer characteristic data of the phase change region and the temperature field acquisition data, a radial basis function interpolation algorithm is input to perform conjugate gradient iterative calculation to obtain temperature field reconstruction data; Inputting the temperature field reconstruction data into a temperature field evolution prediction model to perform temperature field evolution prediction to obtain temperature field prediction data; A dual-objective optimization function including heat exchange efficiency and energy consumption is established according to the temperature field prediction data, and the heat exchanger control parameters are obtained by iterative solution through a particle swarm algorithm.

2. The temperature field analysis method of the heat exchange module according to claim 1, characterized in that: The multi-source data acquisition of the microchannel dual-phase precision phase change heat transfer module is performed to obtain temperature field acquisition data, pressure data, flow data and phase change material characteristic data, including: Performing data acquisition processing on the sensor array of the microchannel dual-phase precision phase change thermal module to obtain original temperature field data, and performing spatial interpolation processing on the temperature sensor measurement signal according to the original temperature field data to obtain temperature field acquisition data; Sampling and processing the inlet and outlet pressure sensor data of the microchannel two-phase precision phase change heat exchange module to obtain pressure data, and sampling and processing the mass flow meter data of the microchannel two-phase precision phase change heat exchange module to obtain flow data; Performing phase change state identification calculation based on the thermocouple array temperature measurement data of the phase change energy storage chamber inside the microchannel dual-phase precision phase change thermal module to obtain phase change completion data, and performing characteristic analysis on the phase change process based on the phase change completion data to obtain phase change start time data and phase change duration data; Substituting the phase change start time data and the phase change duration data into a phase change energy storage calculation model to obtain heat storage capacity data, and performing characteristic calculation on the phase change material according to the heat storage capacity data to obtain phase change material characteristic data.

3. The temperature field analysis method of the heat exchange module according to claim 2, characterized in that: The dimensionless number analysis and calculation of the pressure data and the flow data to obtain a flow pattern criterion value, and the gas-liquid two-phase flow heat transfer coefficient distribution data is calculated according to the flow pattern criterion value, including: Calculate the Reynolds number, Weber number and Kapitza number according to the pressure data and the flow data to obtain dimensionless data of flow characteristics in the microchannel; Performing interval judgment on the dimensionless data of flow characteristics in the microchannel, when the Weber number is less than 0.5, it is judged as an intermittent slug flow area, when the Weber number is between 0.5 and 2.5, it is judged as an annular flow area, and when the Weber number is greater than 2.5, it is judged as a mist flow area, to obtain flow type area judgment data; Calculating the flow pattern conversion characteristic parameters according to the flow pattern area determination data to obtain the flow pattern criterion value, and calculating the boundary layer thickness of the flow pattern criterion value to obtain the boundary layer characteristic data in the microchannel; According to the boundary layer characteristic data in the microchannel and the flow pattern criterion value, the bubble slip coefficient is calculated for the intermittent elastic flow region to obtain the bubble slip characteristic data, and the annular flow liquid film fluctuation is calculated for the flow pattern criterion value to obtain the liquid film fluctuation characteristic data; The droplet entrainment coefficient of the mist flow area is calculated according to the flow type criterion value to obtain droplet entrainment characteristic data, and the bubble slip characteristic data, the liquid film fluctuation characteristic data and the droplet entrainment characteristic data are substituted into the two-phase flow heat transfer coefficient calculation model to obtain the gas-liquid two-phase flow heat transfer coefficient distribution data.

4. The temperature field analysis method of the heat exchange module according to claim 1, characterized in that: The step of inputting the heat transfer characteristic data of the phase change region and the temperature field acquisition data into a radial basis function interpolation algorithm to perform conjugate gradient iterative calculation to obtain temperature field reconstruction data includes: Constructing a radial basis function matrix and a distance weight matrix according to the temperature field acquisition data and a radial basis function interpolation algorithm to obtain an initial interpolation matrix, and normalizing the heat transfer characteristic data of the phase change region to obtain normalized heat transfer characteristic data; Substitute the normalized heat transfer characteristic data into the initial interpolation matrix to obtain a modified interpolation coefficient matrix, and construct a temperature field reconstruction objective function according to the modified interpolation coefficient matrix. The temperature field reconstruction objective function is: ； in, Reconstruct the objective function for the temperature field, is the radial basis function interpolation matrix; is the temperature field to be reconstructed; is the measured temperature data; is the regularization parameter; is the Laplace operator; Performing gradient calculation on the temperature field reconstruction objective function data to obtain objective function gradient data, and performing conjugate direction update calculation according to the objective function gradient data to obtain search direction data; Substitute the search direction data into the linear search algorithm to obtain the optimal step length data, and iteratively update the temperature field based on the optimal step length data. The iteration is stopped when , and the temperature field reconstruction data is obtained.

5. The temperature field analysis method of the heat exchange module according to claim 4, characterized in that: The step of inputting the temperature field reconstruction data into a temperature field evolution prediction model to perform temperature field evolution prediction to obtain temperature field prediction data comprises: Input the temperature field reconstruction data into the encoding layer of the temperature field evolution prediction model for processing to obtain temperature field feature encoding data, wherein the encoding layer includes three convolution sublayers, the convolution kernel size of the first convolution sublayer is 3×3, the step size is 1, and the number of output channels is 64, the convolution kernel size of the second convolution sublayer is 3×3, the step size is 2, and the number of output channels is 128, and the convolution kernel size of the third convolution sublayer is 3×3, the step size is 2, and the number of output channels is 256; Input the temperature field feature encoding data into the long short-term memory network layer of the temperature field evolution prediction model to extract time series features, wherein the long short-term memory network layer includes four parallel memory units, each memory unit includes an input gate, a forget gate and an output gate, and the number of hidden layer neurons is 512, to obtain time series feature data; Performing energy conservation constraint calculation on the time series feature data to obtain constrained time series feature data, and inputting the constrained time series feature data into the attention mechanism layer of the temperature field evolution prediction model for spatial correlation analysis to obtain spatial attention feature data, wherein the attention mechanism layer adopts a self-attention structure, the number of attention heads is 8, and the hidden layer dimension is 256; The spatial attention feature data is input into the decoding layer of the temperature field evolution prediction model to reconstruct the temperature field to obtain temperature field prediction data. The decoding layer includes three deconvolution sublayers. The convolution kernel size of the first deconvolution sublayer is 4×4, the step size is 2, and the number of output channels is 128. The convolution kernel size of the second deconvolution sublayer is 4×4, the step size is 2, and the number of output channels is 64. The convolution kernel size of the third deconvolution sublayer is 3×3, the step size is 1, and the number of output channels is 1.

6. The temperature field analysis method of the heat exchange module according to claim 5, characterized in that: The dual-objective optimization function including heat exchange efficiency and energy consumption is established according to the temperature field prediction data, and the heat exchanger control parameters are obtained by iteratively solving the particle swarm algorithm, including: Construct a first optimization objective function according to the temperature field prediction data And the second optimization objective function ,in: represents the first optimization objective function, is the inlet temperature; is the outlet temperature; represents the second optimization objective function, is the ambient reference temperature; is the fluid density; is the volume flow rate; is the system pressure drop; The bi-objective optimization function is obtained by weighted summation method. ,in: , are the weight coefficients of heat transfer efficiency and energy consumption, respectively, and ; Based on the dual-objective optimization function, the first type of particles are initialized and the second type of particles are initialized to obtain the initial particle group data of the double-layer structure; Inputting the initial particle swarm data of the double-layer structure into a hierarchical fitness calculation model to obtain two layers of fitness, wherein the first layer of the hierarchical fitness calculation model calculates heat exchange efficiency; and the second layer calculates energy consumption; The fitness of the two layers is normalized by a minimum-maximum normalization method to obtain comprehensive fitness data, and non-dominated sorting is performed according to the comprehensive fitness data to obtain hierarchical optimal position data; Performing adaptive speed update calculation on the hierarchical optimal position data to obtain adaptive particle speed data, and performing chaotic position update calculation based on the adaptive particle speed data, performing boundary constraints through mirror mapping, and obtaining target particle position data; A multi-criteria convergence judgment is performed on the target particle position data to obtain multi-dimensional optimal solution set data, and a TOPSIS decision matrix is ​​constructed according to the multi-dimensional optimal solution set data, and the heat exchanger control parameters are calculated by the entropy weight method.

7. A temperature field analysis device for a heat exchange module, characterized in that: A method for analyzing a temperature field of a heat exchange module according to any one of claims 1 to 6, comprising: The acquisition module is used to collect multi-source data of the microchannel dual-phase precision phase change heat transfer module to obtain temperature field acquisition data, pressure data, flow data and phase change material characteristic data; An analysis module, used for performing dimensionless number analysis and calculation on the pressure data and the flow data to obtain a flow pattern criterion value, and calculating gas-liquid two-phase flow heat transfer coefficient distribution data according to the flow pattern criterion value; A calculation module, used for substituting the gas-liquid two-phase flow heat transfer coefficient distribution data and the phase change material characteristic data into the phase change heat transfer control equation for coupling calculation to obtain the phase change region heat transfer characteristic data; An iteration module, used for inputting a radial basis function interpolation algorithm to perform conjugate gradient iterative calculation according to the heat transfer characteristic data of the phase change region and the temperature field acquisition data, so as to obtain temperature field reconstruction data; A prediction module, used for inputting the temperature field reconstruction data into a temperature field evolution prediction model to perform temperature field evolution prediction and obtain temperature field prediction data; The solution module is used to establish a dual-objective optimization function including heat exchange efficiency and energy consumption according to the temperature field prediction data, and obtain the heat exchanger control parameters through iterative solution of the particle swarm algorithm.

8. A temperature field analysis device for a heat exchange module, characterized in that: The temperature field analysis device of the heat exchange module comprises: a memory and at least one processor, wherein the memory stores instructions; The at least one processor calls the instruction in the memory to enable the temperature field analysis device of the heat exchange module to execute the temperature field analysis method of the heat exchange module according to any one of claims 1-6.

9. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instruction is executed by the processor, the temperature field analysis method of the heat exchange module according to any one of claims 1 to 6 is implemented.

Images