Electric control cabinet temperature field simulation modeling method and system

Abstract

The invention relates to the technical field of simulation modeling, in particular to an electric control cabinet temperature field simulation modeling method and system. The method comprises the following steps: firstly, carrying out basic simulation modeling on the electric control cabinet; relevant dynamic parameters and temperature values are set in the temperature sensor; fitting according to the dynamic parameter sequence and the temperature value of each node to obtain a fitting coefficient, and obtaining a feature importance score and an out-of-bag error based on a random forest algorithm; the three are combined to obtain dynamic sensitive intensity; based on the p values of one node and all the rest nodes, the transfer entropy of the temperature, the spatial distance and the dynamic sensitivity intensity, obtaining a coupling heat conduction factor of the node; and the predicted temperature of the next moment is obtained through the temperature sequence, and the temperature of the current moment is combined to obtain the corrected temperature by using the coupling conduction factor weighting, so that the real-time simulation of the temperature field is completed. The precision and adaptability of temperature field simulation of the electric control cabinet are remarkably improved, and the problem of working condition deviation caused by static hypothesis of a traditional simulation model is solved.

Description

Technical Field

[0001] This application relates to the technical field of simulation modeling, and particularly relates to a method and system for simulating and modeling the temperature field of an electrical control cabinet. Background Art

[0002] Electrical control cabinets are used to centrally place electrical components and control equipment, and to achieve functions such as starting, stopping, protecting, and monitoring equipment such as motors. Their structural design needs to consider issues such as heat dissipation and electromagnetic compatibility to ensure the stable operation of the system. Temperature field simulation modeling uses computer technology to establish a mathematical model and perform numerical simulation based on information such as the geometric structure, material properties, and heat source distribution of the electrical control cabinet, in order to predict the temperature distribution inside the electrical control cabinet. Through simulation modeling, potential heat dissipation problems can be discovered in advance during the design stage, the heat dissipation scheme can be optimized, the number of tests and costs in actual production can be reduced, the product R & D cycle can be shortened, and an efficient and low-cost means is provided for the analysis and optimization of complex systems.

[0003] During the actual operation of the electrical control cabinet, dynamic factors such as load fluctuations and changes in ambient temperature and humidity will cause the internal temperature field to change rapidly. Due to the lack of a real-time data correction mechanism in traditional simulation models, it is difficult to accurately reflect the actual working conditions. For example, load fluctuations will cause changes in the heating power of the components inside the electrical control cabinet, and changes in ambient temperature and humidity will affect the heat dissipation efficiency. The combined effect of these factors makes the traditional simulation model have a large error in predicting the internal temperature field of the electrical control cabinet, resulting in design deviations and operation and maintenance risks. Summary of the Invention

[0004] In order to solve the technical problem of large design deviations in the temperature field caused by the lack of real-time data correction, this application provides a method and system for simulating and modeling the temperature field of an electrical control cabinet. The specific technical solutions adopted are as follows: In the first aspect, this application proposes a method for simulating and modeling the temperature field of an electrical control cabinet, which includes the following steps: Model the electrical control cabinet through modeling software, and perform model simulation based on the simulation parameters of the electrical control cabinet to obtain a basic simulation model; Generate random values of dynamic parameters within a fixed range at each moment in the basic simulation model, and form respective sequences of all data of each dynamic parameter up to the current moment; After discretizing the modeling, use the heat conduction principle to obtain the temperature value of each node at each moment, and form a temperature sequence; The dynamic parameters include ambient temperature, ambient humidity, and gas flow rate; Respectively fit the each dynamic parameter sequence and the temperature sequence of each node into a fitting equation to obtain its fitting coefficient; Use the random forest algorithm for all dynamic parameters to respectively obtain their feature importance scores and out-of-bag errors; Based on all the fitting coefficients, feature importance scores of the dynamic parameters and the out-of-bag errors, obtain the dynamic sensitivity intensity; Input the temperature sequences of any two nodes, and output the p-value based on the test algorithm; calculate the transfer entropy of the temperature sequences of the two nodes; calculate the coupling heat conduction factor of a node by combining the mean value of the p-value, spatial distance, and corresponding transfer entropy of the node with the remaining all nodes and the dynamic sensitivity coefficient of the node. Predict the predicted temperature of the node at the next moment through the model based on the temperature sequence; use the normalized coupling heat conduction factor as the weight to weight the temperature of the node at the current moment and the predicted temperature at the next moment to obtain the corrected temperature at the next moment; replace the temperature of the temperature field with the corrected temperature of all nodes at the next moment to realize the real-time simulation of the temperature field in a dynamic environment.

[0005] In the above solution, the present application integrates the multiple linear regression and random forest algorithms, quantifies the linear influence and non-linear pattern of environmental parameters on a single node, constructs a conduction network through transfer entropy and Granger causality test, reveals the directionality and intensity of heat conduction between nodes, and solves the limitations of traditional methods in the fuzzy identification of sensitive nodes and the dependence on empirical assumptions for conduction paths; combines the ARIMA prediction correction model to achieve the real-time iterative optimization of the temperature field. By dynamically weighting and fusing the current temperature and the predicted temperature, the cumulative errors caused by grid discretization and idealization of material properties in traditional finite element simulation are effectively compensated. That is, through dynamic modeling and multi-dimensional data fusion correction mechanism, the accuracy and adaptability of the temperature field simulation of the electric control cabinet are significantly improved, and the problem of working condition deviation caused by static assumptions in traditional simulation models is solved.

[0006] In one embodiment, the simulation parameters include the thermal conductivity, surface radiation emissivity, heat source position and heat source power, and surface convection coefficient.

[0007] In one embodiment, the value range of the environmental temperature is 21~25°C, the value range of the environmental humidity is 45~55%, and the value range of the air flow velocity is 0.5~2 m / s.

[0008] In one embodiment, the method of fitting the dynamic parameter sequence and temperature sequence of each node into a fitting equation to obtain its fitting coefficient is as follows: Set the environmental temperature, environmental humidity, and gas flow velocity of each node as independent variables, and the temperature sequence as the dependent variable, output the fitting equation through the multi-source linear regression algorithm, and obtain the fitting coefficient of each independent variable.

[0009] In one embodiment, the dynamic sensitivity intensity is positively correlated with the feature importance score and the fitting coefficient, and negatively correlated with the out-of-bag error.

[0010] In one embodiment, the method of inputting the temperature sequences of any two nodes and outputting the p-value based on the test algorithm is as follows: Denote any one node as the target point, and use the temperature sequences of the target point and another node as the input; set the null hypothesis that the temperature history of the target point has no predictive ability for the current temperature of the other node, and the alternative hypothesis that the temperature history of the target point has predictive ability for the current temperature of the other node; use the Granger causality test algorithm to output the p-value between the two.

[0011] In one embodiment, the coupling heat conduction factor is negatively correlated with the p-value and the spatial distance, and positively correlated with the transfer entropy and the dynamic sensitivity intensity.

[0012] In one embodiment, the expression of the coupling heat conduction factor is: , represents the p-value between the i-th node and the j-th node, represents the spatial distance between the i-th node and the j-th node, represents the transfer entropy between the temperature sequences of the i-th node and the j-th node, represents the exponential function with the natural constant as the base, N represents the number of nodes, represents the dynamic sensitivity intensity of the i-th node, represents the coupling heat conduction factor of the i-th node.

[0013] In one embodiment, the method of using the normalized coupling heat conduction factor as a weight to weight the temperature of the node at the current moment and the predicted temperature at the next moment to obtain the corrected temperature at the next moment is: , represents the temperature of the i-th node at the current moment, represents the predicted temperature of the i-th node at the next moment, represents the coupling heat conduction factor of the i-th node, represents the corrected temperature of the i-th node at the next moment, is a normalization function.

[0014] In a second aspect, an embodiment of the present application further provides an electric control cabinet temperature field simulation and modeling system, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of the above-mentioned electric control cabinet temperature field simulation and modeling method are implemented.

[0015] The beneficial effects of the present application are: This application integrates multiple linear regression and random forest algorithms, quantifies the linear and non-linear effects of environmental parameters on single nodes, constructs a conduction network through transfer entropy and Granger causality tests, reveals the directionality and intensity of heat conduction between nodes, and solves the limitations of traditional methods in the fuzzy identification of sensitive nodes and the dependence on empirical assumptions for conduction paths. Combining with the ARIMA prediction correction model, it realizes the real-time iterative optimization of the temperature field. By dynamically weighting and fusing the current temperature and the predicted temperature, it effectively compensates for the cumulative errors caused by mesh discretization and idealized material properties in traditional finite element simulations. That is, through dynamic modeling and multi-dimensional data fusion correction mechanisms, it significantly improves the accuracy and adaptability of the temperature field simulation of the electrical control cabinet, and solves the problem of working condition deviation caused by static assumptions in traditional simulation models. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] To more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0017] Figure 1 It is a flowchart of a method for simulating and modeling the temperature field of an electrical control cabinet provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] To further elaborate on the technical means and effects adopted by the present application to achieve the intended invention purpose, the following, in conjunction with the drawings and preferred embodiments, details the specific implementation manners, structures, features, and effects of a method and system for simulating and modeling the temperature field of an electrical control cabinet proposed according to the present application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs.

[0020] An embodiment of a method and system for simulating and modeling the temperature field of an electrical control cabinet: The following specifically describes the specific solutions of a method and system for simulating and modeling the temperature field of an electrical control cabinet provided by the present application with reference to the drawings.

[0021] Please refer to Figure 1 , which shows a flowchart of a method for simulating and modeling the temperature field of an electrical control cabinet provided by an embodiment of the present application. The method includes the following steps: Step S001, perform basic simulation modeling on the electric control cabinet.

[0022] First, use 3D modeling software to create a 3D model based on the design drawing of the electric control cabinet, and simplify secondary details, such as equivalent complex components to regular geometric bodies, ignore screws, small holes, etc., and retain key heat dissipation features. The 3D modeling software used can be SolidWorks for modeling.

[0023] Perform mesh division on the established 3D model to discretize the model into several small subdomains. It should be noted that the type and size of the mesh are determined according to the complexity of the model and the requirements of simulation accuracy. In this embodiment, tetrahedral meshes are used to perform mesh division on the geometric model.

[0024] Set the simulation parameters of the electric control cabinet in the modeling to simulate the model, including the high thermal conductivity of the metal cabinet and the surface radiation emissivity of the components; in this embodiment, the aluminum electric control cabinet used has a thermal conductivity of 237 W / m·K; its surface radiation emissivity is 0.8; in addition, it is also necessary to simulate the heat source position and its power inside the electric control cabinet; the heat sources mainly include the losses of electrical components, such as resistance losses, inductance losses, and switching losses of switching devices, etc. Calculate the heat according to the electrical parameters and working conditions of the components, and set it as the corresponding heat source term, and distribute the heat source to the corresponding surface, so as to obtain the heat generated by each heat source; then, considering the heat dissipation method of the electric control cabinet and the surrounding environmental conditions, set the corresponding boundary conditions. Common boundary conditions include convective heat transfer boundary conditions, radiative heat transfer boundary conditions, and adiabatic boundary conditions, etc. In this embodiment, convective heat transfer boundary conditions are adopted, and the convective coefficient on the cabinet surface is determined to be 3 - 10 W / m²·K.

[0025] Thus, obtain the basic simulation modeling of the temperature field of the electric control cabinet.

[0026] Step S002, set relevant dynamic parameters and temperature values in the basic simulation modeling.

[0027] Traditional simulations need to assume that the load, ambient temperature, humidity, and air flow velocity are constant. However, in actual operation, electric control cabinets often face dynamic interferences such as sudden load changes (such as motor start and stop), day-night temperature differences, humidity fluctuations, and fan speed changes, resulting in significant time-varying characteristics of the temperature field. Ignoring these dynamic factors will underestimate the local high-temperature risk or overestimate the heat dissipation efficiency. Through dynamic modeling, the thermal inertia effect, temperature response delay, and multi-parameter coupling effects can be accurately captured, thereby optimizing the heat dissipation design; The dynamic load is achieved through random power injection, and its fluctuations follow a normal distribution. Based on the central limit theorem, the load power varies randomly around the calibrated value, which is the rated power. A random value within the range of mean ± 3σ is generated for each time step, and the heat source power is updated in real time through the built-in function of the simulation software. In this embodiment, the built-in function used is the randmnormal function of COMSOL, and for a 10kW frequency converter load, the power fluctuates randomly between 8.5kW and 11.5kW.

[0028] Using the same generation method as the dynamic load fluctuations, random temperature, humidity, and air flow velocity are generated. Specifically, to ensure that the electrical control cabinet is in a good ideal working environment, the calibrated value ranges of the ambient temperature and humidity are 21 - 25°C and 45 - 55% respectively, and the specific values in this embodiment are 25°C and 50%; similarly, to ensure the heat dissipation requirements inside the electrical control cabinet and the rotational speed change of the fan, the value range of the air flow velocity is 0.5 - 2 m / s, and the specific value in this embodiment is 1 m / s.

[0029] Define a fixed time step in the transient simulation settings; in this embodiment, the time step takes an empirical value of 0.1 s, that is, the load, temperature, humidity, and air flow velocity all generate random values within the range at a time step of 0.1 s. Use the field variable monitoring function of the simulation software to synchronously record the instantaneous values of the load power, ambient temperature, ambient humidity, and air flow velocity at each time step; and associate a unified time stamp. Subsequently, the load power, ambient temperature, ambient humidity, and air flow velocity recorded from the initial acquisition time to the current time are subjected to overall normalization processing to obtain the load power sequence, ambient temperature sequence, ambient humidity sequence, and gas flow velocity sequence.

[0030] Based on the basic principle of heat conduction, comprehensively considering the heat generated by the heat source, the thermal conductivity of the material, and the heat conduction process in space, establish a heat conduction equation describing the heat transfer inside the electrical control cabinet. Subsequently, using the finite element analysis method, the complex geometric model is discretized into numerous small elements, and each node in the model is used as the basic point for solving the temperature. At each time step, according to the change of each heat source power, numerically solve the heat conduction equation through the finite element analysis software to obtain the temperature value of each node at that moment. As the simulation time progresses, continuously repeat the above solution process to finally obtain the temperature distribution of each node inside the electrical control cabinet at each moment, thereby obtaining the temperature field.

[0031] For each node, the temperature values recorded from the initial acquisition time to the current time of this node form the temperature sequence of this node.

[0032] Thus far, the load power sequence, ambient temperature sequence, ambient humidity sequence, gas flow velocity sequence, and the temperature sequence of each node have been obtained.

[0033] Step S003, obtain the fitting coefficient according to the dynamic parameter sequence and temperature value of each node, and obtain the feature importance score and out-of-bag error based on the random forest algorithm; the three are combined to obtain the dynamic sensitivity strength.

[0034] For each node, changes in data such as ambient temperature, ambient humidity, and airflow velocity will have a significant impact on its temperature field. When the ambient temperature rises, the node heat dissipation becomes more difficult and the temperature rises accordingly. When the ambient temperature decreases, it helps the node to dissipate heat and lowers its temperature. Humidity changes affect the thermal conductivity of the air. When the humidity increases, the thermal conductivity of the air deteriorates, and the node heat dissipation efficiency decreases. When the temperature rises, the humidity decreases, which is conducive to heat dissipation and the node temperature decreases. Changes in airflow velocity change the intensity of convective heat transfer. When the airflow velocity increases, the convective heat transfer is enhanced, the node heat dissipation efficiency is improved, and when the temperature decreases, the airflow velocity decreases, the heat dissipation efficiency decreases, and the node temperature increases. The combined effect of these data makes the node temperature field present a complex dynamic change characteristic.

[0035] Based on the above analysis, taking node i as an example, the ambient temperature sequence, ambient humidity sequence, air flow velocity sequence and temperature sequence of node i are taken as input, the ambient temperature, ambient humidity and gas flow velocity are set as independent variables, and the temperature is set as the dependent variable, and the fitting equation is obtained by fitting. In this embodiment, the multivariate linear regression algorithm is used to output the fitting equation, and the fitting coefficients of each independent variable are obtained, which are respectively recorded as The ambient temperature, ambient humidity and air flow velocity are uniformly recorded as dynamic parameters; then the ambient temperature sequence, ambient humidity sequence and air flow velocity sequence are used as input, and the feature importance score of each dynamic parameter is obtained using the random forest algorithm, which are recorded as , and the out-of-bag error a is obtained through the steps in the random forest algorithm. The fitting coefficients express the linear influence of each dynamic parameter on the node temperature, while the importance score reflects the nonlinear influence of different dynamic parameters on the node temperature. The out-of-bag error represents the prediction error of the random forest model on samples that did not participate in the training.

[0036] The dynamic sensitivity strength of each node is calculated by the feature importance score, fitting coefficient and out-of-bag error of different dynamic parameters; the dynamic sensitivity strength is positively correlated with the feature importance score and fitting coefficient, and negatively correlated with the out-of-bag error.

[0037] It should be noted that positive correlation means that when one variable increases, the other variable also increases, and the two variables change in the same direction. When one variable changes from large to small or from small to large, the other variable also changes from large to small or from small to large. The specific relationship is determined by actual application and this application does not impose any special restrictions.

[0038] It should be noted that negative correlation means that when one variable increases, the other variable decreases accordingly. The change directions of the two variables are opposite. When one variable changes from large to small or from small to large, the other variable also changes from small to large or from large to small. The specific relationship is determined by actual applications, and this application does not make special restrictions.

[0039] Preferably, in this embodiment, the expression of the dynamic sensitivity intensity is: , represents the fitting coefficient of the nth dynamic parameter, represents the characteristic importance score of the nth dynamic parameter, represents the out-of-bag error in the random forest algorithm, represents the exponential function with the natural constant as the base, represents the temperature dynamic sensitivity intensity of the node.

[0040] Among them, the out-of-bag error represents the prediction error of the model on the samples not participating in the training, reflecting the generalization ability of the model to the relationship between dynamic environment parameters and the temperature field. The larger its value, the more complex non-linear relationship or noise interference that the association between environmental parameters and node temperature is not captured by the model, and the less reliable the subsequent calculation of the sensitivity of node temperature to environmental parameters. And the greater the linear influence intensity and non-linear influence intensity of the dynamic parameter, the stronger the contribution of the dynamic parameter to the dynamic sensitivity of the node temperature, and the more significant the influence of the node temperature on environmental fluctuations.

[0041] The dynamic sensitivity characterizes the comprehensive sensitivity intensity of the temperature field of the node to the dynamic parameter. The larger its value, the more easily the temperature of the node is affected by environmental fluctuations, and local overheating or unstable heat dissipation may occur due to dynamic load, temperature and humidity changes, or air flow speed fluctuations.

[0042] So far, the dynamic sensitivity intensity of each node has been obtained.

[0043] Step S004, obtain the coupling heat conduction factor of a node based on the p-value, transfer entropy of temperature, spatial distance, and dynamic sensitivity intensity between the node and all other nodes.

[0044] In the temperature field simulation of the electrical control cabinet, heat is transferred between nodes through heat conduction. When the temperature of adjacent nodes is relatively high, heat will be transferred to the nodes with lower temperature, causing the temperature of these nodes to rise; conversely, if the temperature of adjacent nodes is relatively low, they will absorb the heat of this node, resulting in a decrease in its temperature. This conduction effect does not exist in isolation, but is the result of the combined action of multiple nodes. For example, inside the electrical control cabinet, the nodes close to the heat source will have their temperatures increased due to the heat generated by the heat source, while the nodes far from the heat source balance their temperatures through heat conduction with the surrounding nodes. At the same time, the thermal conductivity of different materials will also affect the conduction effect between nodes. Materials with high thermal conductivity will accelerate the heat transfer between nodes, making the temperature distribution more uniform or change rapidly. In addition, the geometric position of the nodes and the surrounding environmental conditions will also have an impact on the conduction effect. For example, nodes located in positions with better heat dissipation will affect the temperature fields of themselves and surrounding nodes through conduction with other nodes and heat dissipation methods such as convection and radiation with the environment.

[0045] Based on the above analysis, any one node is denoted as the target point, and the temperature sequences of the target point and any other node except the target point are used as inputs, and the p-value is output based on the test algorithm. Preferably, in this embodiment, the null hypothesis is set as that the historical temperature value of the target point has no predictive ability for the current temperature of any other node except the target point, and the alternative hypothesis is that the historical temperature value of the target point has predictive ability for the current temperature of any other node except the target point. No predictive ability indicates no conduction path, and having predictive ability indicates the existence of a conduction path. The Granger causality test algorithm is used to output the p-value between the two.

[0046] Calculate the spatial distance between any two nodes. In this embodiment, the Euclidean distance is used, and the temperature sequences of the two nodes are used as inputs to calculate the transfer entropy between the temperature sequences of the two nodes.

[0047] According to the p-values, spatial distances, and transfer entropies between the target point and all other nodes except the target point, combined with the dynamic sensitivity intensity of the target point, the coupled heat conduction factor of the target point is obtained.

[0048] The coupled heat conduction factor is negatively correlated with the p-value and spatial distance, and positively correlated with the transfer entropy and dynamic sensitivity intensity.

[0049] Preferably, in this embodiment, the expression of the coupled heat conduction factor of the target point is: , represents the p-value between the i-th node and the j-th node, represents the spatial distance between the i-th node and the j-th node, represents the transfer entropy between the temperature sequences of the i-th node and the j-th node, represents the exponential function with the natural constant as the base, and N represents the number of nodes. represents the dynamic sensitivity intensity of the \(i\)-th node represents the coupling heat conduction factor of the \(i\)-th node; the \(i\)-th node is the target point.

[0050] Among them, the dynamic sensitivity intensity is used to measure the sensitivity of the temperature fluctuation of the node to the change of environmental parameters. The larger its value, the stronger the sensitivity of the node itself, and its conduction coupling effect is amplified; the \(p\) value measures the strength of the heat conduction path corresponding to the temperatures of two nodes. The larger its value, the weaker the significance of the existence of the heat path conduction; the spatial distance reflects the spatial proximity of two nodes. The farther the two nodes are, the less likely heat conduction is to occur; the transfer entropy quantifies the amount of information transfer between two nodes, reflecting the direction and strength of heat conduction between nodes. The larger its value, the stronger the heat conduction information flow, that is, when the non-target point is the heat source, its temperature fluctuation will affect the temperature of the target point through the conduction path, avoiding the confusion of bidirectional conduction.

[0051] The larger the value of the coupling conduction factor, the more it indicates that the node is both a sensitive node, vulnerable to environmental fluctuations; and a conduction hub, and the temperature fluctuation diffuses through the conduction path, indicating that the node is more likely to be a node close to the heat source and connected to multiple high heat conduction paths, and is more likely to cause cascading temperature fluctuations due to load mutations.

[0052] So far, the coupling heat conduction factors of each node have been obtained.

[0053] Step S005, obtain the predicted temperature at the next moment through the temperature sequence, and use the coupling conduction factor to weight and obtain the corrected temperature in combination with the temperature at the current moment, so as to complete the real-time simulation of the temperature field.

[0054] Normalize the coupling heat conduction factors of all nodes. In this embodiment, the softmax function is used for normalization, so that the sum of the coupling heat conduction factors of all nodes at the current moment is 1. Input the temperature sequence of the nodes, use the ARIMA model to predict the predicted temperature of the nodes at the next moment, and use the coupling heat conduction factors of the nodes as weights to weight the predicted temperature at the next moment and the temperature at the current moment to obtain the corrected temperature at the next moment.

[0055] Preferably, in this embodiment, the expression of the corrected temperature is: , represents the temperature of the \(i\)-th node at the current moment, represents the predicted temperature of the \(i\)-th node at the next moment, represents the coupling heat conduction factor of the \(i\)-th node, represents the corrected temperature of the \(i\)-th node at the next moment, is the normalization function.

[0056] Calculate the corrected temperature at the next moment according to the time change, and replace the temperature in the current temperature field with it to realize the real-time simulation of the temperature field in a dynamic environment.

[0057] Based on the same inventive concept as the above method, an embodiment of the present invention further provides an electric control cabinet temperature field simulation and modeling system, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of any one of the above methods for simulating and modeling the temperature field of an electric control cabinet are implemented.

[0058] It should be noted that 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 foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

[0059] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. The key point of each embodiment is to illustrate the differences from other embodiments.

Claims

1. A method for simulating and modeling the temperature field of an electric control cabinet, characterized in that: The method comprises the following steps: Model the electric control cabinet through modeling software, and simulate the model based on the simulation parameters of the electric control cabinet to obtain basic simulation modeling; In basic simulation modeling, random values ​​of dynamic parameters in a fixed range are generated at each moment, and all data of each dynamic parameter before the current moment form their own sequence; after discretizing the modeling, the temperature value of each node at each moment is obtained using the principle of heat conduction to form a temperature sequence; dynamic parameters include ambient temperature, ambient humidity and gas flow rate; Fit each dynamic parameter sequence and temperature sequence of each node into a fitting equation to obtain its fitting coefficient; obtain the feature importance score and out-of-bag error of all dynamic parameters through the random forest algorithm; obtain the dynamic sensitivity intensity based on all fitting coefficients, feature importance scores and out-of-bag errors of dynamic parameters; Input the temperature series of any two nodes and output the p value based on the test algorithm; calculate the transfer entropy of the temperature series of the two nodes; calculate the coupled heat conduction factor of the node by combining the p value, spatial distance, corresponding transfer entropy of a node with all other nodes and the dynamic sensitivity coefficient of the node; Based on the temperature sequence, the model is used to predict the predicted temperature of the node at the next moment; the normalized coupled heat conduction factor is used as the weight to weight the temperature of the node at the current moment and the predicted temperature at the next moment to obtain the corrected temperature at the next moment; the corrected temperature of all nodes at the next moment replaces the temperature field temperature to realize real-time simulation of the temperature field in a dynamic environment.

2. The method for simulating and modeling the temperature field of an electric control cabinet according to claim 1, characterized in that: The simulation parameters include thermal conductivity, surface radiation emissivity, heat source position and heat source power, and surface convection coefficient.

3. The method for simulating and modeling the temperature field of an electric control cabinet according to claim 1, characterized in that: The ambient temperature ranges from 21 to 25° C., the ambient humidity ranges from 45 to 55%, and the air flow velocity ranges from 0.5 to 2 m / s.

4. The method for simulating and modeling the temperature field of an electric control cabinet according to claim 1, characterized in that: The method of fitting each dynamic parameter sequence and temperature sequence of each node into a fitting equation to obtain its fitting coefficient is: The ambient temperature, ambient humidity and gas flow rate of each node are set as independent variables, and the temperature series is set as the dependent variable. The fitting equation is output through the multi-source linear regression algorithm to obtain the fitting coefficient of each independent variable.

5. The method for simulating and modeling the temperature field of an electric control cabinet according to claim 1, characterized in that: The dynamic sensitivity strength is positively correlated with the feature importance score and the fitting coefficient, and negatively correlated with the out-of-bag error.

6. The method for simulating and modeling the temperature field of an electric control cabinet according to claim 1, characterized in that: The method of inputting the temperature series of any two nodes and outputting the p value based on the test algorithm is: Record any node as the target point, and take the temperature series of the target point and another node as input; set the null hypothesis that the temperature history value of the target point has no predictive ability for the current temperature of another node, and the alternative hypothesis that the temperature history value of the target point has predictive ability for the current temperature of another node; use the Granger causality test algorithm to output the p-value between the two.

7. The method for simulating and modeling the temperature field of an electric control cabinet according to claim 1, characterized in that: The coupled heat conduction factor is negatively correlated with the p value and the spatial distance, and is positively correlated with the transfer entropy and the dynamic sensitivity intensity.

8. The method for simulating and modeling the temperature field of an electric control cabinet according to claim 7, characterized in that: The expression of the coupled heat conduction factor is: , represents the p-value of the i-th node and the j-th node, represents the spatial distance between the i-th node and the j-th node, represents the transfer entropy between the temperature series of the i-th node and the j-th node, represents an exponential function with a natural constant as the base, N represents the number of nodes, represents the dynamic sensitivity strength of the i-th node, Represents the coupled heat conduction factor of the i-th node.

9. The method for simulating and modeling the temperature field of an electric control cabinet according to claim 1, characterized in that: The method of using the normalized coupled heat conduction factor as a weight to weight the temperature of the node at the current moment and the predicted temperature at the next moment to obtain the corrected temperature at the next moment is: , represents the temperature of the i-th node at the current moment, represents the predicted temperature of the i-th node at the next moment, represents the coupled heat conduction factor of the ith node, represents the corrected temperature of the i-th node at the next moment, is the normalization function.

10. A temperature field simulation modeling system for an electric control cabinet, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the electric control cabinet temperature field simulation modeling method as described in any one of claims 1-9 are implemented.

Images