Method for optimizing thermal performance of substation enclosure based on cfd and energy consumption simulation

Abstract

This invention discloses a method for optimizing the thermal performance of substation building envelopes based on CFD and energy consumption simulation, belonging to the field of substation building energy conservation technology. This invention addresses the problem of excessive CFD simulation examples when constructing a year-round meteorological parameter-near-wall temperature prediction model by identifying typical representative meteorological parameters using a clustering algorithm. It also solves the coupling problem between CFD simulation and energy consumption simulation by constructing a meteorological parameter-near-wall temperature prediction model based on SVM regression, which involves hourly simulation steps. This invention couples CFD and energy consumption simulation, addressing the current problem of not considering the impact of high-heat sources on the building envelope and energy consumption in energy consumption simulations. Thus, this invention improves the accuracy of substation building envelope thermal performance evaluation and optimization.

Description

Technical Field

[0001] This invention relates to the field of energy-saving technology for substation buildings, specifically a method for optimizing the thermal performance of substation building envelopes based on CFD and energy consumption simulation. Background Technology

[0002] The building envelope not only serves to divide the interior space of the building, but also acts as a medium for heat exchange between the building space and the indoor and outdoor environments. For substation buildings, different rooms correspond to different internal heat sources and temperature control requirements. Among them, high-heat-generating equipment rooms such as main transformer rooms and reactor rooms (heat generation ≥1000kW) rely on natural ventilation for temperature control; air-conditioned rooms such as secondary equipment rooms and power distribution rooms require air conditioning to maintain the equipment in an operating environment of 26-28℃.

[0003] Different rooms in a substation require different thermal performance parameters for their building envelope. For example, insulation is prioritized in summer, heat dissipation in transitional seasons, and thermal insulation may be necessary in winter. Furthermore, the intensity of solar radiation received varies depending on the substation site. These factors necessitate a trade-off between heat dissipation and thermal insulation in the building envelope's thermal performance, posing a challenge to the life-cycle carbon reduction optimization design of the building envelope.

[0004] Due to the unique characteristics of substation buildings, especially those with high-heat sources (such as transformer radiators) nearby, near-wall temperature simulations using Computational Fluid Dynamics (CFD) are necessary to assess the impact of these high-heat sources on the building's energy consumption and envelope. However, to match the hourly simulation step accuracy of energy consumption simulations, CFD requires hourly calculations of the entire year's operating conditions, resulting in significant computational resource consumption and high time costs. Therefore, current energy consumption simulations for optimizing the thermal performance of substation envelopes cannot accurately assess the impact of external high-heat sources on the envelope. Furthermore, current energy consumption simulations for optimizing the thermal performance of substation envelopes do not consider heat exchange between external high-heat sources and the envelope due to convection and radiation. This leads to an underestimation of the surface temperature of the envelope during the simulation, resulting in a significant discrepancy with actual conditions. In summary, current methods for evaluating and optimizing the thermal performance of substation envelopes suffer from insufficient accuracy. Summary of the Invention

[0005] The purpose of this invention is to address the problems in existing methods for optimizing the thermal performance of substation envelopes, namely, the inability to accurately assess the impact of external high-heat sources on the envelope, and the failure to consider heat exchange between external high-heat sources and the envelope due to convective and radiative heat transfer. This leads to insufficient accuracy in the evaluation and optimization of substation envelope thermal performance. The invention provides a method for optimizing the thermal performance of substation envelopes based on CFD and energy consumption simulation. It uses a clustering algorithm to select typical representative meteorological parameters for CFD simulation to address the problem of excessive simulation examples. It constructs a meteorological parameter-near-wall temperature prediction model based on SVM regression to solve the coupling problem between CFD and energy consumption simulations at hourly simulation steps. Furthermore, it couples CFD and energy consumption simulations to address the current problem of not considering the impact of high-heat sources on the external envelope and building energy consumption during energy consumption simulations, thereby improving the accuracy of substation envelope thermal performance evaluation and optimization.

[0006] The objective of this invention is mainly achieved through the following technical solutions:

[0007] A method for optimizing the thermal performance of substation building envelope based on CFD and energy consumption simulation includes the following steps:

[0008] Step S1: Obtain hourly meteorological parameters for the location of the target substation throughout the year;

[0009] Step S2: Use the K-Means clustering algorithm to perform cluster analysis on the meteorological parameters and extract K groups of meteorological parameters;

[0010] Step S3: Construct a CFD analysis model of the target substation;

[0011] Step S4: Use the CFD analysis model of the target substation to simulate the external wall temperature value of each set of meteorological parameters under the influence of the heat source;

[0012] Step S5: Based on meteorological parameters and corresponding outer wall temperature values, SVM regression is performed using the radial basis function as the kernel function to obtain the outer wall temperature prediction model.

[0013] Step S6: Based on the epw file of the target substation location, construct an hourly external wall temperature dataset for the whole year using the external wall temperature prediction model;

[0014] Step S7: Input the outer wall temperature dataset as the boundary condition into the energy consumption simulation software, and combine it with the genetic optimization algorithm to optimize the thermal performance of the building envelope.

[0015] Furthermore, the meteorological parameters obtained in step S1 include four parameters: air temperature, wind speed, wind direction, and solar radiation intensity.

[0016] Furthermore, after obtaining the meteorological parameters in step S1, the process also includes preprocessing the obtained meteorological parameters. The preprocessing includes removing duplicate data, default values, and outliers, as well as performing Min-Max normalization.

[0017] Furthermore, step S2 includes the following steps:

[0018] Step S21: Use the elbow rule and profile coefficient as dual indicators to evaluate and test the optimal value of the cluster number K.

[0019] Step S22: Execute the K-Means clustering algorithm, using the centroid of each cluster as the typical moment meteorological parameter, and output K sets of meteorological parameters.

[0020] Furthermore, in step S3, when constructing the CFD analysis model of the target substation, the mesh is divided in the following manner:

[0021] The mesh orthogonality quality is greater than 0.8, the boundary layer mesh has 10 to 15 layers, and the first layer mesh height is... satisfy , The calculation formula is as follows:

[0022]

[0023] in, The normal phase distance is the distance between the wall and the phase. For fluid density, The characteristic velocity of the flow. For dynamic viscosity, This represents the wall shear stress.

[0024] Furthermore, step S1 also includes obtaining the hourly heat output of the radiators at the target substation location throughout the year; step S4 specifically includes the following steps: inputting K sets of meteorological parameters into the CFD analysis model of the target substation, using the heat output of the radiators as boundary conditions, conducting CFD analysis model simulation iteration calculations to obtain K sets of external wall temperature field data, and using the area-weighted average temperature as the external wall temperature value for each set of external wall temperature field data; wherein, the area-weighted average temperature... The calculation formula is as follows:

[0025]

[0026] in, T n No. n Temperature values ​​of each grid surface A n No. n The area of ​​each grid face. N The total number of mesh faces on the selected surface.

[0027] Furthermore, after obtaining the outer wall temperature value in step S4, outliers are removed and the original data is normalized. Then, the meteorological parameters and the corresponding outer wall temperature value data are divided into a training set and a test set according to a set ratio. In step S5, the outer wall temperature prediction model is obtained by training based on the meteorological parameters and the corresponding outer wall temperature values ​​in the training set.

[0028] Furthermore, step S5 also includes using R... 2 As an evaluation index for overfitting diagnosis, R0 2 Calculate using the following formula:

[0029]

[0030] in, This is a predicted value for the outer wall surface temperature. It is the input sample value of the outer wall surface temperature. R is the average value of the input outer wall surface temperature sample values. 2 The closer to 1, the better the fit; conversely, the closer to 0, the worse the fit.

[0031] Furthermore, step S7 includes the following steps:

[0032] Step S71: Construct a three-dimensional model of the target substation, build an EnergyPlus energy consumption simulation model based on the three-dimensional model of the target substation, and calculate the operating energy consumption of air conditioning and fans throughout the entire life cycle of the target substation using the EnergyPlus energy consumption simulation model.

[0033] Step S72: In the Grasshopper platform, use the Honeybee energy consumption simulation plugin, input the target substation enclosure structure construction information, room disturbances, timetable and temperature control system parameters, generate an EnergyPlus executable idf file, which contains boundary conditions for the target substation energy consumption calculation, and call the EnergyPlus energy consumption simulation model to calculate the energy consumption of the target substation during operation.

[0034] Step S73: Based on the Grasshopper platform, operate on the 3D model of the target substation, select the enclosure structure adjacent to the high heat source object and extract the name of the enclosure structure, change the outer boundary condition type of the enclosure structure to temperature boundary condition in the idf file, and enter the annual hourly outer wall temperature dataset from step S6 in the outer boundary condition parameter field.

[0035] Step S74: Call the Galapagos genetic algorithm plugin built into the Grasshopper platform, use the thickness of the insulation layer in the air-conditioned room envelope as the optimization variable, and use the goal of minimizing the annual operating energy consumption of the target substation as the objective to perform iterative optimization.

[0036] Furthermore, in the method for optimizing the thermal performance of substation building envelope based on CFD and energy consumption simulation, the parameters of the genetic algorithm in step S74 are set as follows:

[0037] Initial population size: 50-100 individuals;

[0038] Crossover rate: 60%-80%;

[0039] Variation rate: 5%-10%;

[0040] Convergence criteria: Energy consumption improvement of less than 1% for 20 consecutive generations or reaching the upper limit of 100 generations of iteration.

[0041] This invention uses a clustering algorithm to screen and classify typical representative meteorological types throughout the year; for each typical representative meteorological type, CFD simulation of the external wall temperature is used to obtain a meteorological parameter-external wall temperature dataset; based on the meteorological parameter-external wall temperature dataset, an external wall temperature prediction model (SVM regression prediction model) is established to construct an hourly external wall temperature dataset for the whole year; the annual external wall temperature dataset is used as the boundary condition input for the external envelope in energy consumption simulation to carry out substation energy consumption simulation analysis; based on optimization algorithms such as genetic algorithms, with the goal of minimizing energy consumption, the thermal parameters of the envelope are optimized.

[0042] In summary, compared with existing technologies, this invention has the following advantages: To consider the impact of high heat sources on the building envelope, this invention couples CFD and energy consumption simulation techniques. Furthermore, it uses a clustering algorithm and an SVM-based external wall temperature prediction model, thus overcoming the technical bottleneck of inconsistent simulation step size accuracy when CFD and energy consumption simulation are coupled, which leads to an excessive number of CFD cases, huge computational resource consumption, and high time costs. Unlike conventional substation building envelope thermal performance optimization, this invention can consider the load and energy consumption impact of high heat sources near the building envelope, as well as the influence of solar radiation intensity. This makes the energy consumption simulation more realistic, the optimization effect on the building envelope thermal performance more accurate, and thus improves the accuracy of substation building envelope thermal performance evaluation and optimization. Attached Figure Description

[0043] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings:

[0044] Figure 1A flowchart of a specific embodiment of the present invention;

[0045] Figure 2 This is a flowchart illustrating the operational logic of a specific embodiment of the present invention. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of this invention are only for explaining this invention and are not intended to limit this invention.

[0047] Example:

[0048] like Figure 1 and Figure 2 As shown, the method for optimizing the thermal performance of substation envelope based on CFD and energy consumption simulation includes the following steps: Step S1, obtaining hourly meteorological parameters for the target substation location throughout the year; Step S2, using the K-Means clustering algorithm to perform cluster analysis on the meteorological parameters and extract K groups of meteorological parameters; Step S3, constructing a CFD analysis model for the target substation; Step S4, using the CFD analysis model for the target substation to simulate the external wall temperature value of each group of meteorological parameters under the influence of heat sources; Step S5, based on the meteorological parameters and corresponding external wall temperature values, using the radial basis function kernel function as the kernel function to perform SVM regression to obtain an external wall temperature prediction model; Step S6, based on the epw file of the target substation location, using the external wall temperature prediction model to construct an hourly external wall temperature dataset for the whole year; Step S7, using the external wall temperature dataset as the boundary condition input to the energy consumption simulation software, and combining it with a genetic optimization algorithm to optimize the thermal performance of the envelope.

[0049] The meteorological parameters obtained in step S1 of this embodiment include four parameters: air temperature, wind speed, wind direction, and solar radiation intensity. After obtaining the meteorological parameters in step S1, this embodiment also includes preprocessing the obtained meteorological parameters. Preprocessing includes removing duplicate data, default values, and outliers, and using Min-Max normalization to eliminate the influence of dimensions between parameters. In this embodiment, duplicate data refers to records with identical timestamps. When processing default values, for data with consecutive default values ​​≤3 hours, linear interpolation between adjacent timestamps is used to fill in the gaps; for values ​​exceeding 3 hours, all data for the day is deleted. When processing outliers, a threshold is set based on physical feasibility (e.g., temperature ∈ [-40, 50]℃, wind speed ∈ [0, 40]m / s); values ​​outside this range are considered outliers and removed. The formula for calculating Min-Max normalization is:

[0050]

[0051] in, X These are the original parameter values. 𝑋 min This is the minimum value of the parameter throughout the year. 𝑋 max This is the maximum value of the parameter throughout the year. X norm The normalized parameter values ​​(range [0,1]).

[0052] In this embodiment, the wind direction is a circular angle (0°~360°). Direct normalization would result in a discontinuity between 0° and 360°. Converting to a unit vector:

[0053]

[0054] Thus, the final input parameters for clustering in this embodiment are: temperature, wind speed, solar radiation intensity, 𝑉 𝑥 , V y A total of 5 dimensions; 𝑉 𝑥 and V y The vector decomposition components of wind direction are used to convert the wind direction angle in a circular direction into a continuous variable in a Cartesian coordinate system, in order to solve the problem of the circular discontinuity of wind direction. 𝑉 𝑥 For the east-west component, V y This represents the north-south component.

[0055] Step S2 of this embodiment includes the following steps: Step S21, using the elbow rule and the silhouette coefficient as dual indicators to evaluate and test the optimal value of the number of clusters K; Step S22, executing the K-Means clustering algorithm, using the centroid of each cluster as the meteorological parameter at a typical moment, and outputting K sets of meteorological parameters.

[0056] In this embodiment, step S2 employs a dual-index collaborative optimization strategy to avoid subjective bias, including the elbow rule:

[0057] Calculate the number of clusters k From 2 to K max (Preferred) The sum of squares (WCSS) within a cluster (N=8760):

[0058]

[0059] by k Plot a graph with the horizontal axis and WCSS as the vertical axis, selecting points of abrupt slope change (elbow points) as candidates. K 1 .

[0060] Profile coefficient:

[0061] Calculate each k Corresponding average profile coefficient S ( k ):

[0062]

[0063] Where a(i) is the average distance from sample i to other points in the same cluster, and b(i) is the average distance from sample i to the nearest other cluster.

[0064] The cluster number corresponding to the maximum value of S(k) is selected as a candidate. K 2 .

[0065] final K Value determined:

[0066] like K 1 and K 2 If they coincide, then that value is used; otherwise, S(k) > 0.5 is preferred. S ( k The clustering quality is considered to be greater than 0.5 (medium clustering quality), and the smaller of WCSS(k) / WCSS(k−1) > 0.85 is taken. k value.

[0067] This embodiment uses K-Means+ to initialize the centroid (to avoid getting trapped in local optima by random initialization), and uses Euclidean distance as the metric to iterate until the centroid change is <0.1% or the upper limit of 300 iterations is reached.

[0068] Typical parameter generation: for each cluster C i Calculate the centroid 𝜇 𝑖 Denormalize the normalized centroid parameters:

[0069]

[0070] And the wind direction is determined by the vector ( 𝑉 𝑥 , V y Restored to angle:

[0071]

[0072] The results are converted to a range of 0° to 360°. Finally, K sets of typical meteorological parameters are generated.

[0073] In step S3 of this embodiment, when constructing the CFD analysis model of the target substation, 3D modeling software such as Rhino and SpaceClaim is used. A 3D geometric model of the substation is established based on the substation floor plan, and 3D geometric models of high-heat-generating objects such as transformer radiators and reactor radiators are established based on the electrical equipment drawings. These models are combined to form a complete CFD analysis model of the target substation. When constructing the CFD analysis model of the target substation, it is necessary to consider the building envelope (walls, doors, windows, roof, etc.), high-heat-generating sources such as transformer / reactor radiators, ventilation structures such as louvered openings, underground facilities such as cable trenches and ventilation ducts, and equipment layout such as the relative positions of radiators and the building envelope. The use of 3D modeling software such as Rhino and SpaceClaim to construct the CFD analysis model of the target substation in this embodiment is feasible based on existing technology and will not be elaborated further here.

[0074] This embodiment uses meshing software such as ICEM or Fluent Meshing to generate a CFD analysis model of the target substation, and then imports the mesh into CFD simulation software such as Fluent for simulation. CFD technology is used to simulate the temperature impact of high heat sources such as radiators on the external wall surface of the substation, thereby accurately assessing the energy consumption of the substation's air conditioning and ventilation equipment and optimizing the thermal performance of the external wall. To accurately assess the external wall surface temperature, the overall mesh orthogonality quality should be >0.8 during mesh generation, and the mesh should be refined specifically for the substation's external wall surface, with 10-15 boundary layer mesh layers, while ensuring that the height of the first layer mesh meets the requirements. , The calculation formula is as follows:

[0075]

[0076] in, The normal phase distance is the distance between the wall and the phase. For fluid density, The characteristic velocity of the flow. For dynamic viscosity, This refers to the wall shear stress. In this embodiment, a requirement is set for the height of the first layer of mesh on the outer wall surface during mesh generation to achieve accurate evaluation. This embodiment imports the target substation CFD analysis model into CFD simulation software such as Fluent for simulation. Based on CFD technology, it simulates the temperature impact on the substation's outer wall surface caused by high-heat-generating objects such as radiators under real-world conditions through convective and radiative heat transfer, thereby optimizing the performance of the target substation CFD analysis model.

[0077] Step S1 of this embodiment also includes obtaining the hourly heat output of the radiators at the target substation location throughout the year. Step S4 of this embodiment specifically includes the following steps: inputting K sets of meteorological parameters into the CFD analysis model of the target substation, using the radiator heat output as the boundary condition, and conducting iterative simulation calculations of the CFD analysis model to obtain K sets of external wall surface temperature field data. Because the obtained external wall surface temperature field data exhibits uneven distribution on the external wall surface, this embodiment uses the area-weighted average temperature as the external wall surface temperature value for each set of external wall surface temperature field data. The core idea is that larger areas contribute more to the overall average value, thus more accurately reflecting the actual physical state. Among them, the area-weighted average temperature... The calculation formula is as follows:

[0078]

[0079] in, T n No. n Temperature values ​​of each grid surface A n No. n The area of ​​each grid face. N The total number of mesh faces on the selected surface.

[0080] In step S4 of this embodiment, after obtaining the outer wall temperature value, outliers are removed and the original data is normalized. Then, the meteorological parameters and corresponding outer wall temperature values ​​are divided into a training set and a test set according to a set ratio. In step S5 of this embodiment, the outer wall temperature prediction model is obtained by training based on the meteorological parameters and corresponding outer wall temperature values ​​in the training set. In specific implementation of this embodiment, the data is divided into a training set and a test set at a ratio of 3:1. The test set is used to verify the generalization ability of the outer wall temperature prediction model.

[0081] Step S5 in this embodiment also includes using R to predict the external wall temperature model. 2 As an evaluation index for overfitting diagnosis, R0 2 Calculate using the following formula:

[0082]

[0083] in, This is a predicted value for the outer wall surface temperature. It is the input sample value of the outer wall surface temperature. R is the average value of the input outer wall surface temperature sample values. 2 The closer to 1, the better the fit; conversely, the closer to 0, the worse the fit.

[0084] Step S7 of this embodiment includes the following steps: Step S71: Construct a three-dimensional model of the target substation, construct an EnergyPlus energy consumption simulation model based on the three-dimensional model of the target substation, and calculate the operating energy consumption of the air conditioner and fan throughout the entire life cycle of the target substation using the EnergyPlus energy consumption simulation model; Step S72: Use the Honeybee energy consumption simulation plugin in the Grasshopper platform, input the structural information of the target substation enclosure, room disturbances, timetable and temperature control system parameters, generate an executable EnergyPlus idf file, which contains boundary conditions for calculating the energy consumption of the target substation, and call the EnergyPlus energy consumption simulation model to calculate the energy consumption of the target substation during operation; Step S73: Operate on the three-dimensional model of the target substation based on the Grasshopper platform, select the enclosure structure adjacent to the high heat source object and extract the enclosure structure name, change the outer boundary condition type of the enclosure structure to temperature boundary condition in the idf file, and input the annual hourly outer wall temperature dataset from step S6 in the outer boundary condition parameter field. Step S74: Call the Galapagos genetic algorithm plugin built into the Grasshopper platform, using the insulation layer thickness in the air-conditioned room envelope as the optimization variable, and aiming to minimize the annual operating energy consumption of the target substation for iterative optimization. Before implementing step S7 in this embodiment, the architectural plan data of the target substation is collected to obtain design information for energy consumption simulation, including building floor plans, envelope construction information, room disturbances (electrical equipment loss heat value, lighting power density), timetables (electrical equipment loss change timetable, lighting power change timetable, room thermostat timetable, ventilation system start-stop timetable), and temperature control system parameters (air conditioning system cooling capacity, cooling energy efficiency, ventilation system air volume, power consumption per unit air volume). In step S71 of this embodiment, based on the target substation architectural floor plan, a 3D model of the target substation that can be applied to EnergyPlus is built using Rhino and Grasshopper. The operating energy consumption value obtained in step S71 is used to evaluate the impact of high heat source objects on operating energy consumption and as an evaluation value for optimizing the thermal performance of the envelope.In step S72 of this embodiment, the input of the building envelope construction information is used to model and accurately reflect the thermal performance parameters of the building envelope. The insulation layer thickness is set as a variable parameter, and its value range is defined through the Number Slider module. During subsequent energy consumption simulation, it can dynamically respond to changes in the insulation layer thickness parameter and output annual energy consumption data. The input of room disturbance is used to accurately reflect the amount of heat generated in the building. The input of equipment loss change timetable and lighting power change timetable is used to simulate the changes in indoor heat generation during the actual operation of the substation. The input of room thermostat timetable and ventilation system start-stop timetable is used to simulate the start-stop status of the air conditioning system and ventilation system during substation operation. The input of temperature control system parameters (air conditioning system cooling capacity, cooling energy efficiency, ventilation system air volume, power consumption per unit air volume) is used to accurately model the air conditioning and ventilation system, help to accurately calculate the operating energy consumption of the air conditioning and ventilation system, and finally generate an IDF file. In this embodiment, step S73 involves manually selecting the enclosure structure adjacent to the high-heat-generating object, obtaining the name of the selected enclosure structure, and locating the enclosure structure character segment whose boundary condition type needs to be modified in the IDF file by searching for the enclosure structure name. The outer boundary condition type of the enclosure structure is then changed to a temperature boundary condition. This embodiment uses the annual outer wall temperature dataset as a boundary condition in energy consumption calculations. In this embodiment, step S74, when specifically implemented, utilizes the Grasshopper platform and calls the built-in Galapagos genetic algorithm plugin to optimize the thermal performance of the building envelope. The insulation layer thickness is used as a genetic variable, connected to the Genome input of the Galapagos plugin. The energy consumption simulation result output is connected to the Fitness input of Galapagos, and the optimization direction is set to minimize energy consumption. The initial population size is set to 50-100 individuals to balance computational efficiency and global search capability. The crossover rate is set to 60%-80% to promote superior gene combinations, and the mutation rate is set to 5%-10% to avoid getting trapped in local optima. Optimization iterations are performed, with a maximum of 100 iterations to avoid infinite computation. Convergence is defined as an improvement of less than 1% for 20 consecutive generations.

[0085] The operating logic of this embodiment is as follows:

[0086] 1. Based on the K-Means clustering method, hourly meteorological conditions throughout the year are clustered for analysis. The meteorological data includes four variables: air temperature, wind speed, wind direction, and solar radiation intensity. This algorithm can identify typical representative weather parameters reflecting the weather conditions at different times of the year. CFD simulations based on these typical weather parameters significantly reduce the number of CFD simulation examples. Furthermore, the typical weather parameter set obtained through K-means clustering covers various meteorological models throughout the year, avoiding data redundancy and improving the training efficiency of subsequent prediction models. This effectively reduces the time cost of constructing a yearly external wall temperature dataset.

[0087] 2. Based on SVM regression, a predictive model is constructed and trained using typical meteorological parameters and their corresponding CFD simulation results for external wall temperature. Using this predictive model, external wall temperature values ​​for different time periods and meteorological parameters throughout the year can be quickly constructed, achieving the requirement of hourly simulation steps in energy consumption simulation with a relatively small number of CFD cases.

[0088] 3. The external wall temperature dataset is input as a boundary condition into the energy consumption simulation software, and then the thermal performance of the building envelope is optimized using an optimization algorithm. This method solves the problem of time accuracy mismatch between CFD and energy consumption simulation, and can thus consider the impact of high heat sources near the building envelope on the building envelope, making the substation energy consumption simulation more realistic and the thermal performance optimization results more accurate.

[0089] This embodiment addresses the impact of high-heat sources on the thermal performance of building envelopes and the accuracy of energy consumption simulation in special substation scenarios. It proposes a systematic solution integrating meteorological clustering, CFD sample calculation, SVM regression prediction, custom input of energy consumption simulation boundaries, and optimization algorithms. It efficiently constructs a prediction model of "meteorological parameters -> external wall temperature considering high-heat sources," and accurately uses the prediction results as key boundary conditions for energy consumption simulation, thus achieving high-precision coupling and optimization at an acceptable computational cost.

[0090] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing the thermal performance of substation building envelope based on CFD and energy consumption simulation, characterized in that, Includes the following steps: Step S1: Obtain hourly meteorological parameters for the location of the target substation throughout the year; Step S2: Use the K-Means clustering algorithm to perform cluster analysis on the meteorological parameters and extract K groups of meteorological parameters; Step S3: Construct a CFD analysis model of the target substation; Step S4: Use the CFD analysis model of the target substation to simulate the external wall temperature value of each set of meteorological parameters under the influence of the heat source; Step S5: Based on meteorological parameters and corresponding outer wall temperature values, SVM regression is performed using the radial basis function as the kernel function to obtain the outer wall temperature prediction model. Step S6: Based on the epw file of the target substation location, construct an hourly external wall temperature dataset for the whole year using the external wall temperature prediction model; Step S7: Input the outer wall temperature dataset as the boundary condition into the energy consumption simulation software, and combine it with the genetic optimization algorithm to optimize the thermal performance of the building envelope.

2. The method for optimizing the thermal performance of substation envelope based on CFD and energy consumption simulation according to claim 1, characterized in that, The meteorological parameters obtained in step S1 include four parameters: air temperature, wind speed, wind direction, and solar radiation intensity.

3. The method for optimizing the thermal performance of substation building envelope based on CFD and energy consumption simulation according to claim 1, characterized in that, After obtaining the meteorological parameters in step S1, the method further includes preprocessing the obtained meteorological parameters. The preprocessing includes removing duplicate data, default values, and outliers, as well as using Min-Max normalization.

4. The method for optimizing the thermal performance of substation envelope based on CFD and energy consumption simulation according to claim 1, characterized in that, Step S2 includes the following steps: Step S21: Use the elbow rule and profile coefficient as dual indicators to evaluate and test the optimal value of the cluster number K. Step S22: Execute the K-Means clustering algorithm, using the centroid of each cluster as the typical moment meteorological parameter, and output K sets of meteorological parameters.

5. The method for optimizing the thermal performance of substation envelope based on CFD and energy consumption simulation according to claim 1, characterized in that, When constructing the CFD analysis model of the target substation in step S3, the mesh is divided in the following manner: The mesh orthogonality quality is greater than 0.8, the boundary layer mesh has 10 to 15 layers, and the first layer mesh height is... satisfy , The calculation formula is as follows: ； in, The normal phase distance of the wall. For fluid density, The characteristic velocity of the flow. For dynamic viscosity, This represents the wall shear stress.

6. The method for optimizing the thermal performance of substation envelope based on CFD and energy consumption simulation according to claim 1, characterized in that, Step S1 further includes obtaining the hourly heat output of the radiators at the target substation location throughout the year; Step S4 specifically includes the following steps: inputting K sets of meteorological parameters into the CFD analysis model of the target substation, using the heat output of the radiators as boundary conditions, conducting CFD analysis model simulation iteration calculations to obtain K sets of external wall temperature field data, and using the area-weighted average temperature as the external wall temperature value for each set of external wall temperature field data; wherein, the area-weighted average temperature... The calculation formula is as follows: ； in, T n No. n Temperature values ​​of each grid surface A n No. n The area of ​​each grid face. N The total number of mesh faces on the selected surface.

7. The method for optimizing the thermal performance of substation enclosure structures based on CFD and energy consumption simulation according to claim 1, characterized in that, After obtaining the outer wall temperature value in step S4, outliers are removed and the original data is normalized. Then, the meteorological parameters and the corresponding outer wall temperature values ​​are divided into a training set and a test set according to a set ratio. In step S5, the outer wall temperature prediction model is obtained by training based on the meteorological parameters and the corresponding outer wall temperature values ​​in the training set.

8. The method for optimizing the thermal performance of substation building envelope based on CFD and energy consumption simulation according to claim 1, characterized in that, Step S5 also includes using R to predict the external wall temperature. 2 As an evaluation index for overfitting diagnosis, R0 2 Calculate using the following formula: ； in, This is a predicted value for the outer wall surface temperature. It is the input sample value of the outer wall surface temperature. R is the average value of the input outer wall surface temperature sample values. 2 The closer to 1, the better the fit; conversely, the closer to 0, the worse the fit.

9. The method for optimizing the thermal performance of substation enclosure structures based on CFD and energy consumption simulation according to any one of claims 1 to 8, characterized in that, Step S7 includes the following steps: Step S71: Construct a three-dimensional model of the target substation, build an EnergyPlus energy consumption simulation model based on the three-dimensional model of the target substation, and calculate the operating energy consumption of air conditioning and fans throughout the entire life cycle of the target substation using the EnergyPlus energy consumption simulation model. Step S72: In the Grasshopper platform, use the Honeybee energy consumption simulation plugin, input the target substation enclosure structure construction information, room disturbances, timetable and temperature control system parameters, generate an EnergyPlus executable idf file, which contains boundary conditions for the target substation energy consumption calculation, and call the EnergyPlus energy consumption simulation model to calculate the energy consumption of the target substation during operation. Step S73: Based on the Grasshopper platform, operate on the 3D model of the target substation, select the enclosure structure adjacent to the high heat source object and extract the name of the enclosure structure, change the outer boundary condition type of the enclosure structure to temperature boundary condition in the idf file, and enter the annual hourly outer wall temperature dataset from step S6 in the outer boundary condition parameter field. Step S74: Call the Galapagos genetic algorithm plugin built into the Grasshopper platform, use the thickness of the insulation layer in the air-conditioned room envelope as the optimization variable, and use the goal of minimizing the annual operating energy consumption of the target substation as the objective to perform iterative optimization.

10. The method for optimizing the thermal performance of substation envelope based on CFD and energy consumption simulation according to claim 9, characterized in that, The parameters of the genetic algorithm in step S74 are set as follows: Initial population size: 50-100 individuals; Crossover rate: 60%-80%; Variation rate: 5%-10%; Convergence criteria: Energy consumption improvement of less than 1% for 20 consecutive generations or reaching the upper limit of 100 generations of iteration.

Images