Ultra-low energy consumption building optimization design method
By constructing a dataset and using support vector machines based on engineering priorities to improve the model and multi-objective optimization algorithm, the problems of low design efficiency and insufficient parameter optimization in ultra-low energy building design are solved, achieving more efficient energy consumption and cost optimization.
Patent Information
- Application Number
- CN202511447025.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-11
- Publication Date
- 2025-11-18
AI Technical Summary
Existing ultra-low energy building design methods rely on the designer's experience, resulting in low design efficiency, lack of parameter optimization, and deficiencies in comprehensively judging economic efficiency and applicability, especially in dealing with the relationship between visible light transmittance of windows and shading.
A dataset was constructed using energy consumption analysis software and engineering cost estimation software. An improved support vector machine model based on engineering priorities was established, and combined with an improved multi-objective optimization algorithm (NSGA-II), the design parameters were optimized through nonlinear correlation to achieve the optimal solution for energy consumption and cost.
It improves design efficiency, reduces the frequency of designers relying on experience, provides a basis for parameter optimization, enhances the economy and applicability of ultra-low energy buildings, and optimizes design parameters to achieve lower energy consumption and construction costs.
Smart Images

Figure CN120974608A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of energy-saving building, and particularly relates to an ultra-low energy consumption building optimization design method. BACKGROUND
[0002] With the rapid development of economic society, the environmental problems and energy crisis in the global range have become the main challenges to restrict the sustainable development of society. As an important component of national economy, the energy consumption and carbon emission of the construction industry have become the core problems concerned in the field of energy saving. In recent years, with the development of building energy saving technology, the ultra-low energy consumption building has been widely applied in the field of construction engineering due to its good energy saving effect.
[0003] From the existing literature and design cases at home and abroad, the ultra-low energy consumption building mainly adopts the "index design method", and the design process mainly includes the following steps: firstly, the design parameters are initially selected according to the requirements of the existing building energy saving specification of the building scheme; secondly, the energy consumption analysis software is applied to modeling, and the energy consumption of the designed building is simulated; thirdly, the simulation analysis result is judged, when the heat transfer coefficient of the envelope structure and the comprehensive energy consumption meet the specification requirements of the ultra-low energy consumption building, the building can be identified as the ultra-low energy consumption building; fourthly, when the heat transfer coefficient of the envelope structure or the comprehensive energy consumption index does not meet the specification requirements of the ultra-low energy consumption building, the first step needs to be restarted.
[0004] The ultra-low energy consumption building design method adopting the "index design method" has the following defects: firstly, the selection of design parameters more depends on the experience of designers, and when the designers lack experience, the design efficiency is low due to multiple parameter selection; secondly, the ultra-low energy consumption building designed by adopting the "index design method" meets the technical requirements of the ultra-low energy consumption building, but lacks the parameter optimization process, and has great defects in the comprehensiveness and applicability, such as how to deal with the contradictory relationship between the visible light transmittance of the window body and the sunshade; thirdly, in the scheme stage of design, the parameters of multiple professions are still in the initial stage, so that the selection of key parameters of the ultra-low energy consumption building lacks basis. Therefore, the selection of appropriate design parameters with the constraint indexes of comprehensive energy consumption and cost is an important basis for judging the design quality of the ultra-low energy consumption building. SUMMARY
[0005] In view of the defects of the prior art, the main purpose of the present application is to provide an ultra-low energy consumption building optimization design method to solve one or more problems in the prior art.
[0006] The technical scheme of the present application is as follows:
[0007] An ultra-low energy consumption building optimization design method, comprising the following steps:
[0008] obtaining the design parameters of the ultra-low energy consumption building;
[0009] With the design parameters as input, the energy consumption of the ultra-low energy consumption building under different design parameters is obtained by using energy consumption analysis software, and the cost of the ultra-low energy consumption building under different design parameters is obtained by using engineering cost software, and the energy consumption dataset and the cost dataset are constructed correspondingly;
[0010] An improved support vector machine model based on engineering priority is constructed;
[0011] Based on the constructed energy consumption dataset and cost dataset, the improved support vector machine model is trained to obtain the nonlinear correlation between the design parameters and the energy consumption, and the nonlinear correlation between the design parameters and the cost;
[0012] An improved multi-objective optimization algorithm based on engineering priority is determined;
[0013] With energy consumption and cost as constraints, the improved multi-objective optimization algorithm is used to perform multi-objective optimization on the nonlinear correlation to obtain an optimal solution.
[0014] The beneficial effects of the present application relative to the prior art at least include:
[0015] (1) For a designed building, by transforming the parameters affecting the design of the ultra-low energy consumption building, a large number of datasets are obtained by using energy consumption analysis software and engineering cost calculation software, thereby solving the problem of low design efficiency caused by manual selection of parameters by designers based on experience.
[0016] (2) An engineering priority setting method for samples in the dataset is proposed, an improved support vector machine improved model based on engineering priority is proposed, and by training a small number of samples, the nonlinear correlation between the ultra-low energy consumption design parameters and the energy consumption and cost is established, thereby reducing the frequency of calling the energy consumption analysis software and the engineering cost software in the design process.
[0017] (3) The engineering priority of the design parameters of the ultra-low energy consumption building is determined, and an improved multi-objective optimization algorithm (NSGA-II) based on the engineering priority of the design parameters is proposed, which improves the engineering implementability of multi-objective optimization and provides a basis for the initial selection of parameters in the scheme stage.
[0018] It should be understood that the content described in the summary section is not intended to limit the key or important features of the embodiments of the present application, nor to limit the scope of the present application. Other features of the present application will become apparent from the following description. In addition, the implementation of any embodiment of the present application does not mean that multiple or all of the above beneficial effects are simultaneously possessed or achieved. BRIEF DESCRIPTION OF DRAWINGS
[0019] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following will briefly introduce the drawings required in the embodiments or prior art description. Obviously, the drawings in the following description are only exemplary, and for those skilled in the art, other drawings can be obtained from the provided drawings without creative labor.
[0020] The structures, proportions, sizes, etc. shown in the specification are only used to cooperate with the content disclosed in the specification, for understanding and reading by those skilled in the art, and are not used to limit the conditions for implementing the present application, so they do not have technical significance. Any modification of structure, change of proportion relationship or adjustment of size, without affecting the effects and purposes that can be achieved by the present application, should still fall within the scope of the technical content disclosed by the present application.
[0021] Figure 1 A schematic diagram of the principle of the ultra-low energy consumption building optimization design method of some embodiments of the present application;
[0022] Figure 2 A schematic diagram of the flow of the ultra-low energy consumption building optimization design method of some embodiments of the present application;
[0023] Figure 3 A schematic diagram of the analysis model of an engineering case of the present application;
[0024] Figure 4 A schematic diagram of the comparison of the building energy consumption simulation analysis and prediction results of a specific embodiment of the present application;
[0025] Figure 5 A schematic diagram of the comparison of the building cost simulation analysis and prediction results of a specific embodiment of the present application;
[0026] Figure 6 A schematic diagram of the multi-objective optimization results of a specific embodiment of the present application. DETAILED DESCRIPTION
[0027] In order to make the purpose, technical solutions and advantages of the embodiments of the present application more clear, the following will further describe the embodiments of the present application in combination with the drawings. Here, the schematic embodiments of the present application and their descriptions are used to explain the present application, but not as a limitation of the present application.
[0028] In the present application, unless otherwise explicitly specified and limited, the terms "mounting", "connection", "connecting", "fixing" and the like should be understood in a broad sense, for example, can be fixed connection, can also be detachable connection, or integral; can be mechanical connection, can also be electrical connection; can be directly connected, can also be indirectly connected through an intermediate medium, can be internal communication of two elements or interaction relationship of two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.
[0029] It should be understood that the terms "including", "containing", "consisting of" or any other variant are intended to cover non-exclusive inclusion, so that the product, device, process or method including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such product, device, process or method. Without more limitation, the elements defined by the statement "including", "containing", "consisting of" do not exclude the presence of additional identical elements in the product, device, process or method comprising the elements.
[0030] It should also be understood that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices, components or structures referred to must have a particular orientation, be constructed or operated in a particular orientation, and cannot be understood as a limitation on the present application.
[0031] In addition, the terms "first", "second" are only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more features. In the description of the present application, the meaning of "multiple" is two or more, unless otherwise explicitly specified and limited.
[0032] The implementation of the present application is described in detail below in combination with the preferred embodiments.
[0033] As shown in Figure 1 , Figure 1 The principle block diagram of the method for optimizing the design of an ultra-low energy consumption building is shown. The goal of the optimization design of the ultra-low energy consumption building is to select appropriate design parameters to achieve the optimization of energy consumption and cost, which is essentially a multi-objective optimization problem. The method mainly includes the following steps: 1. data acquisition; 2. model training; 3. multi-objective optimization. The three steps are executed in sequence, and data interaction is performed between them.
[0034] Based on the above principles, as shown in Figure 2 The method comprises the following steps: 101, obtaining design parameters of the super low energy consumption building; 102, inputting the design parameters, obtaining energy consumption of the super low energy consumption building under different design parameters by using energy consumption analysis software, and obtaining construction cost of the super low energy consumption building under different design parameters by using construction cost software, and constructing energy consumption data set and construction cost data set correspondingly; 103, constructing an improved support vector machine model based on engineering priority; 104, training the improved support vector machine model based on the constructed energy consumption data set and construction cost data set, obtaining nonlinear correlation between the design parameters and energy consumption, and nonlinear correlation between the design parameters and construction cost; 105, determining an improved multi-objective optimization algorithm based on engineering priority; 106, performing multi-objective optimization on the nonlinear correlation by using the improved multi-objective optimization algorithm under the constraints of energy consumption and construction cost, and obtaining an optimal solution.
[0035] 101, obtaining design parameters of the super low energy consumption building;
[0036] In this step, according to the energy consumption characteristics and construction cost influencing factors of the super low energy consumption building, the design parameters include two or more of the following: external wall heat transfer coefficient, roof heat transfer coefficient, partition wall heat transfer coefficient, external door and window heat transfer coefficient, and window-wall ratio.
[0037] 102, inputting the design parameters, obtaining energy consumption of the super low energy consumption building under different design parameters by using energy consumption analysis software, and obtaining construction cost of the super low energy consumption building under different design parameters by using construction cost software, and constructing energy consumption data set and construction cost data set correspondingly;
[0038] This link mainly performs energy consumption simulation and construction cost analysis; wherein:
[0039] The energy consumption simulation specifically comprises: inputting the design parameters of the super low energy consumption building, obtaining energy consumption of the super low energy consumption building under different design parameters by using energy consumption analysis software, and constructing multiple energy consumption data sets;
[0040] The construction cost analysis specifically comprises: inputting the design parameters of the super low energy consumption building, obtaining construction cost of the super low energy consumption building under different design parameters by using construction cost software, and constructing multiple construction cost data sets.
[0041] In this step, the energy consumption simulation obtains the energy consumption of the designed building according to the design parameters of the super low energy consumption building. The simulation result reflects the energy consumption level of the building under various combinations of design parameters. The construction cost analysis obtains the construction cost of the super low energy consumption part according to the design parameters of the super low energy consumption building, that is, the investment increased to achieve the goal of the super low energy consumption building. The simulation result reflects the construction cost of the building under various combinations of design parameters.
[0042] The energy consumption and cost of the ultra-low energy consumption building are influenced by multiple factors, and the building envelope has the greatest influence on the energy consumption and cost of the building. The parameters of the building envelope, i.e., the design parameters of the ultra-low energy consumption building, mainly include: the heat transfer coefficient of the outer wall, the heat transfer coefficient of the roof, the heat transfer coefficient of the partition wall, the heat transfer coefficient of the outer door and window, the window-wall ratio, etc.
[0043] The parameters are coupled with each other, and a single parameter cannot play a decisive role in the design quality. In order to obtain as much data as possible, the design parameters of the ultra-low energy consumption building are input in the present application, and the energy consumption and cost results under different design parameters are obtained by using existing energy consumption analysis software (such as PKPM-Energy) and engineering cost software (such as Guanglianda), and multiple, for example, 100 energy consumption data sets and 100 cost data sets are constructed for subsequent model training and verification. In order to ensure the comparability of the data in the data sets, the design parameters, energy consumption and cost of the ultra-low energy consumption building are normalized respectively.
[0044] In some embodiments, constructing the energy consumption data set includes: taking the design parameters of the ultra-low energy consumption building as input, and obtaining the energy consumption of the ultra-low energy consumption building under different design parameters by using the energy consumption analysis software, and constructing the energy consumption data set.
[0045] In the present application, the energy consumption data set is a nonlinear correlation of the input parameters such as the heat transfer coefficient of the outer wall, the heat transfer coefficient of the roof, the heat transfer coefficient of the partition wall, the heat transfer coefficient of the outer door and window, and the window-wall ratio and the energy consumption.
[0046] In the present application, each energy consumption data set contains a group of design parameters and corresponding energy consumption values.
[0047] In some embodiments, constructing the cost data set includes: taking the design parameters of the ultra-low energy consumption building as input, and obtaining the cost of the ultra-low energy consumption building under different design parameters by using the engineering cost software, and constructing the cost data set.
[0048] In the present application, the cost data set is a nonlinear correlation of the input parameters such as the heat transfer coefficient of the outer wall, the heat transfer coefficient of the roof, the heat transfer coefficient of the partition wall, the heat transfer coefficient of the outer door and window, and the window-wall ratio and the cost.
[0049] In the present application, each cost data set contains a group of design parameters and corresponding cost.
[0050] 103, constructing an improved support vector machine model based on engineering priority;
[0051] As is readily appreciated, a support vector machine (SVM) is a supervised learning model used for classification and regression analysis. An SVM finds an optimal hyperplane to separate the sample points into two classes. This optimal hyperplane not only separates the two classes of samples, but also maximizes the distance (i.e., margin) of the two classes from the hyperplane. In this way, the SVM can improve the generalization ability of the model as much as possible while maintaining the classification accuracy.
[0052] The traditional support vector machine (SVM) is based on the statistical theory and the structural risk minimization theory, has the advantages of simple structure and fast calculation speed, and the complexity of the model does not change with the change of the input sample dimension, so it is very suitable for processing high-dimensional problems. However, the traditional SVM still has some limitations in regression analysis, that is, the traditional SVM does not consider the engineering priority of the data in regression analysis, resulting in the averaging of the regression results, and some SVM improvement algorithms consider the sample quality, but do not give the value that meets the practical significance of engineering priority. Based on the above reasons, the present application constructs an improved support vector machine model based on engineering priority. The model fully considers the engineering priority, so that different data have different contributions to the training of the improved support vector machine, and the high-quality data have greater contribution to the model training, which is consistent with the tendency to select the design result in engineering events.
[0053] In some embodiments, constructing an improved support vector machine model based on engineering priority includes:
[0054] (1) Setting the engineering priority: according to the two indexes of energy consumption and cost of ultra-low energy consumption building analysis, the engineering priority IC i,E and IC i,I are defined respectively, and the values are:
[0055] ;
[0056] ;
[0057] In the formula, IC i,E is the engineering priority of the energy consumption data set i, IC i,I is the engineering priority of the cost data set i, the subscript E represents energy consumption, I represents cost, E i is the comprehensive energy consumption of the energy consumption data set i, E b is the comprehensive energy consumption of the reference building, j is the energy consumption reduction step of the design building relative to the reference building, n is the number of the energy consumption data set i, C i is the cost of the cost data set i, C j is the cost of the cost data set j, and m is the number of the cost data set i.
[0058] (2) Constructing an improved model of support vector machine based on engineering priority:
[0059]
[0060] where w is a weight vector, b is a bias term, ξ is a relaxation factor, s is an engineering priority vector of the data set i, s i is the engineering priority of the data set i, T is a transpose identifier, C is a penalty factor, ξ i is the relaxation factor of the data set i, s.t. is a constraint condition, y i is a label function, Φ is a normalization function, x i is the i-th set of design parameters of the data set i.
[0061] In this step, the importance of data in the improved model of support vector machine is reflected by engineering priority.
[0062] In some embodiments, setting the engineering priority includes determining the engineering priority of the k-th design parameter in the data set i, expressed as:
[0063]
[0064] It is easy to understand that the data set i here does not distinguish between the energy consumption data set i or the cost data set i, that is, the same algorithm is used for both the energy consumption data set and the cost data set.
[0065] wherein, is the engineering priority of the k-th design parameter in the data set i, is the impact of a unit change of the k-th design parameter on energy consumption / cost, is the impact of a unit change of the j-th design parameter on energy consumption / cost, n is the number of design parameters, k is the serial number of the design parameter, and j is the energy consumption reduction step of the j-th design parameter relative to the reference building.
[0066] 104, based on the constructed energy consumption data set and cost data set, model training is performed on the improved model of support vector machine to obtain the nonlinear correlation between the design parameters and energy consumption, and the nonlinear correlation between the design parameters and cost;
[0067] The energy consumption analysis software and the cost analysis software are called to perform model training on the improved model of support vector machine using the data set constructed by the early energy consumption simulation and cost analysis, to obtain the nonlinear correlation between the design parameters of the super low energy consumption building and energy consumption and cost.
[0068] In this step, the purpose of model training is to obtain the nonlinear correlation between the design parameters of super low energy consumption buildings and energy consumption and cost. The nonlinear correlation refers to "establishing the correlation between design parameters and energy consumption and cost through the improved model after training, as the input and output of this correlation are not in a linear relationship, it is called nonlinear correlation in mathematics". The internal relationship of nonlinear correlation is that the output parameter can be obtained from the input parameter by establishing nonlinear correlation, and then the input parameter and the output parameter are optimized. For example, for a function y=f(x), the input parameter is x and the output parameter is y, and the extreme value of f(x) can be solved by applying a multi-objective optimization algorithm. In the present application, since the input parameter x and the output parameter y cannot establish a display function relationship, the improved support vector machine is used to establish the nonlinear correlation between x and y.
[0069] In this step, the model training is consistent with the training process of the traditional support vector machine; the nonlinear correlation refers to that the energy consumption and cost can be calculated according to the input parameters by using the improved support vector machine based on engineering priority, and since the input parameters and energy consumption and cost cannot be expressed by a specific formula, they are collectively referred to as nonlinear correlation. The improved support vector machine based on engineering priority is used to establish the nonlinear correlation between the input parameters and energy consumption and cost, which avoids manual repetition and improves the automation degree of model training.
[0070] 105, determining an improved multi-objective optimization algorithm based on engineering priority;
[0071] In this step, the basis of the multi-objective optimization algorithm is NSGA-II, and the basic idea of this algorithm is to increase the selection and sorting of non-dominated solutions and the calculation of crowding distance on the basis of genetic algorithm, which solves the problems of poor optimization effect and slow speed of traditional multi-objective optimization methods in the absence of experience. In the optimization design of super low energy consumption, the engineering priority of design parameters is used to improve the crossover and mutation operation of NSGA-II, so as to make the optimization model consistent with engineering practice.
[0072] In some embodiments, determining an improved multi-objective optimization algorithm based on engineering priority includes:
[0073] Improving the crossover method and / or mutation method of NSGA-II by using the engineering priority of design parameters;
[0074] Obtaining an improved NSGA-II algorithm based on engineering priority.
[0075] As can be easily understood, in the multi-objective optimization algorithm, the process of recalculating energy consumption and cost by calling the improved support vector after the parameters are crossed and mutated is the interaction in step Figure 1 .
[0076] In this step, the crossover is the process of randomly recombining the parent chromosomes in the optimization process, and the present application adopts the two-level crossover method of fusion data engineering priority and parameter engineering priority based on decimal floating point numbers. For any ith group of design parameters x i and jth group of design parameters x j , the crossover condition is:
[0077]
[0078] In the formula, s i is the engineering priority of the ith group of design parameters x i , s j is the engineering priority of the ith group of design parameters x j , and ε is the engineering priority threshold value. When the distance of the engineering priorities of two data x i and x j is greater than ε, x i and x j can be crossed, and the purpose of this setting is to guarantee the genetic advantage of the data with higher engineering priority. When the crossover operation needs to be performed between data with large difference in engineering priority, a larger threshold value ε can be taken in the implementation process.
[0079] It is easy to understand that in terms of energy consumption, s i and s j actually equal IC i,E , but the values are different for different data x i and data x j . Similarly, in terms of cost, s i and s j actually equal IC i,I , but the values are different for different data x i and data x j .
[0080] For any two data x i and x j , when they meet the above crossover condition, the crossover operation can be performed. In order to improve the operation efficiency, the design parameter priority is defined in the present application, and the crossover operation is preferentially performed for the design parameter with high priority. The probability that the kth design parameter in data x i is selected for crossover is defined as:
[0081]
[0082] In the formula, is the engineering priority of the kth design parameter in the ith group of design parameters x i , n is the number of design parameters, The i-th group of design parameters x i The engineering priority of the j-th design parameter in x
[0083] When the k-th parameter of x i and x j is a selected parameter, the k-th parameter of x i and x j is subjected to a crossover operation.
[0084] The engineering priority of the design parameters improves the probability of the parameters that have a greater impact on energy consumption and cost being selected, and avoids the problem of randomness in the original NSGA-II.
[0085] In this step, mutation is a phenomenon in which part of the chromosomes of the data is mutated in the evolution process. In the optimization design of ultra-low energy consumption buildings, new solutions can be obtained through the mutation of design parameters, so that the optimization process gradually approaches the optimal solution. In order to ensure that the mutation operation is consistent with engineering practice, the mutation operation is performed preferentially when the engineering priority of the parameter is high. The design parameter k is selected to perform the mutation operation when the following formula is satisfied:
[0086]
[0087] In the formula, The engineering priority of the k-th design parameter in the i-th group of design parameters x i , n is the number of design parameters, and rnd( ) is a random number with as the benchmark, The engineering priority of the j-th design parameter in the i-th group of design parameters x i , and θ is the design parameter priority threshold.
[0088] 106. The nonlinear correlation is subjected to multi-objective optimization by using the improved multi-objective optimization algorithm, and the optimal solution is obtained.
[0089] In this step, the optimal solution of the energy consumption and cost is obtained under the constraints of the heat transfer coefficient of the external wall, the heat transfer coefficient of the roof, the heat transfer coefficient of the partition wall, the heat transfer coefficient of the external door and window, and the window-wall ratio of the four orientations. The constraint refers to the control of the input parameters to ensure that the energy consumption after optimization is not too high and the cost is also not too high. The optimal solution is obtained by using the improved multi-objective optimization algorithm with energy consumption and cost as constraints. The main steps include:
[0090] S1: The N data generated by the support vector machine improved model are used as the initial data set, which is defined as P0, and the input vector set is , wherein x iFor the i group design parameters of ultra-low energy consumption building, the engineering value is the heat transfer coefficient of external wall, the heat transfer coefficient of roof, the heat transfer coefficient of partition wall, the heat transfer coefficient of external door and window, and the window-wall ratio of four orientations; let P t =P0;
[0091] S2: non-dominated sorting is performed on the data set P t , and the non-dominated sorting of each data p i ∈p t is calculated, and the number of samples n i dominated by it and the dominated data set are calculated; wherein t is the current generation number; the sample is derived from the improved support vector machine algorithm, which is actually the input data and output data of the support vector machine;
[0092] The purpose of non-dominated sorting is to sort the obtained solutions, and the higher the ranking, the better the solution.
[0093] S3: calculate the data crowding degree;
[0094] S4: select data according to non-dominated sorting and crowding degree, and generate offspring data Q t through crossover and mutation operation;
[0095] S5: combine P t and Q t into a joint data set R t , and perform non-dominated sorting on R t ;
[0096] S6: calculate the data crowding degree and crowding distance of R t , and select N data to form a new data set P t+1 ;
[0097] S7: if the maximum evolution generation number has been reached, go to step S8, otherwise, go to step S2;
[0098] S8: output the design parameters corresponding to the optimal energy consumption and cost.
[0099] In S3, the calculation of data crowding degree is as follows:
[0100]
[0101] In the formula, d i is the crowding degree of data p i , and are the function values of adjacent data of data p i in the target η, and are two extreme values of the target η respectively;
[0102] It is easy to understand that the first d i is a new d i , the second d i is the original d i . Similar to the statement i = i + 1 in C language, it means that i + 1 is assigned to i.
[0103] It is easy to understand that the design parameters corresponding to the optimal energy consumption and cost are the two optimal goals achieved under the constraints of energy consumption and cost. In fact, it is to constantly generate new solutions through continuous crossover and mutation, and then find the best one among these solutions. Then, the input parameters (input parameters refer to the heat transfer coefficient of the outer wall, the heat transfer coefficient of the roof, etc.) of the generated energy consumption and cost are extracted as the design parameters of the super low energy consumption building.
[0104] Engineering case test:
[0105] The above optimization method is described in detail with a building in Shanghai as an example.
[0106] The main functional rooms of the building are ordinary classrooms, art classrooms and offices. The first floor has a height of 4.5 meters, the second to fourth floors have a height of 3.9 meters, and the building height is 17.0 meters. The roof insulation uses extruded polystyrene board, and the heat transfer coefficient is not greater than 0.4; the outer wall and the partition wall use 200-400 thick insulation modular blocks, and the heat transfer coefficient is not greater than 0.4; the outer window uses heat-insulating aluminum alloy, and the heat transfer coefficient is not greater than 1.4; the window-wall ratio of the four orientations is not greater than 0.7. The analysis model is shown in Figure 3 .
[0107] The main steps of optimizing the building into a super low energy consumption building are as follows:
[0108] (1) Data acquisition
[0109] The energy consumption of the selected case is mainly concentrated in the air conditioning system and the lighting system, as shown in Tables 1 and 2, which show the design requirements of the case, aiming to optimize the design parameters to meet the design requirements. This project uses split air conditioners, and the outdoor unit is set on the outdoor building facade.
[0110] The indoor setting parameters of the air conditioning system are shown in Table 1:
[0111] Table 1 Air conditioning system parameter setting
[0112]
[0113] The office, ordinary classroom and art classroom of this project all use high-efficiency LED lamps. The lighting power LPD is not higher than 70% of the target value specified in the specification. The lighting system parameters are shown in Table 2:
[0114] Table 2 Lighting system parameter setting
[0115]
[0116] Using PKPM-Energy simulation analysis software and Glodon cost estimation software, 100 samples (datasets) were obtained by varying the heat transfer coefficients of exterior walls, roofs, partition walls, exterior doors and windows, and the window-to-wall ratio for the four orientations. These samples included energy consumption and cost datasets under the same design parameters. Some samples are listed below:
[0117] (0.28, 0.21, 0.9, 1.4, 0.17, 0.40, 0.35, 0.40, 0.82, 0.68);
[0118] (0.35, 0.26, 0.88, 1.3, 0.22, 0.38, 0.29, 0.37, 0.92, 0.59);
[0119] The first 8 data points in the sample are design parameters (input data), the 9th is energy consumption data, and the 10th is cost data.
[0120] (2) Model training
[0121] The improved support vector machine model based on engineering priorities maps data from ultra-low energy buildings from the original input space to a high-dimensional feature space, achieving more accurate and reliable regression analysis. Using 100 samples acquired during the data acquisition phase, with 70 samples as the training set and 30 samples as the test set, the test results of the improved support vector machine are as follows: Figure 4 , Figure 5 As shown. Among them Figure 3 The horizontal axis represents 30 sample test data points, and the vertical axis represents energy consumption values. Figure 4 The simulated data in the model are energy consumption and cost data obtained using energy consumption analysis software; the predicted data are energy consumption data calculated using an improved support vector machine model based on project priorities. Figure 5 The horizontal axis represents 30 sample test data, and the vertical axis represents the cost. Figure 5 The simulated data in the data are cost data obtained using cost analysis software; the predicted data are cost data calculated using an improved support vector machine model based on project priorities.
[0122] Figure 4 , Figure 5 The purpose is to compare the calculation results of the method proposed in this invention with the simulation results using PKPM and Glodon software. When the difference between the two is not large, it indicates that the method proposed in this invention is effective. For example, if the actual result of a certain function (corresponding to the simulation result) is y, and the prediction result obtained by our new method is y1, if the difference between y and y1 is very small, it indicates that our prediction method is effective.
[0123] (3) Multi-objective optimization
[0124] According to the nonlinear correlation between energy consumption and cost obtained by the improved support vector machine model based on engineering priority, the optimal solution set is obtained by the improved multi-objective optimization algorithm based on engineering priority, so that the best input vector is obtained.
[0125] The algorithm uses its fast non-dominated sorting method to retain the optimal individual after each population evolution iteration to the next generation, thereby maintaining the diversity of the population. The NSGA-II multi-objective optimization algorithm applies a binary tournament selection strategy to generate a population of size N, then performs non-dominated sorting on the population, and finally applies a mutation operator evolution operation to obtain the next generation of individuals. The results when the population evolution generation is 600 are shown in Figure 6 . Figure 6 The horizontal axis is the cost, and the vertical axis is the energy consumption. The graph expresses the correlation between energy consumption and cost, which aims to provide a reference for decision makers.
[0126] In this engineering case, the optimal solution (i.e., the lowest energy consumption and cost) is finally obtained, and the corresponding design parameters (such as the external wall heat transfer coefficient) are also given. These design parameters can be provided to designers as parameter selection for super low energy consumption building design, avoiding the selection of parameters based on experience or blindly.
[0127] The present application takes the external wall heat transfer coefficient, the roof heat transfer coefficient, the partition wall heat transfer coefficient, the external door and window heat transfer coefficient, and the four orientation window wall ratios as input parameters, and takes the energy consumption and cost as optimization objectives. An improved support vector machine model considering data engineering priority is proposed, and a nonlinear correlation between input parameters and energy consumption and cost is established. An improved multi-objective optimization algorithm based on input parameter engineering priority is proposed to optimize the design parameters of super low energy consumption buildings.
[0128] More specifically, the following innovations are made:
[0129] First, the engineering priority of data is proposed, and an improved support vector machine model (support vector machine improved model) is designed based on the engineering priority of data, which makes the model training accept more data that meets the needs of engineering practice and has a positive impact on model training. The simulation analysis results prove the accuracy of the improved support vector machine prediction results. By subtracting the vertical coordinate value of the circle point from the vertical coordinate value of the square point and then dividing by the vertical coordinate value of the square point, the maximum error of the predicted energy consumption of the 30 test samples is 9.86%, and the maximum error of the predicted cost is 7.62%.
[0130] Secondly, the engineering priority of the design parameters is proposed, and the improved multi-objective optimization algorithm is proposed, and the optimal solution is better than the existing design result in energy consumption and cost. Specifically, in the test case, the original design result energy consumption result is 0.82, and through the algorithm, the optimized energy consumption is 0.74, and the energy consumption is reduced by 9.8%. The original design result cost result is 0.66, and through the algorithm optimization, it is 0.62, and the cost is reduced by 6.1%. It plays an important role in the design of optimized low-energy consumption buildings.
[0131] Those skilled in the art can easily understand that the above preferred embodiments can be freely combined and superimposed without conflict.
[0132] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application.
Claims
1. A method for optimizing the design of ultra-low energy consumption buildings, characterized in that, Includes the following steps: Obtain design parameters for ultra-low energy buildings; Using design parameters as input, energy consumption analysis software is used to obtain the energy consumption of the ultra-low energy consumption building under different design parameters, and engineering cost software is used to obtain the cost of the ultra-low energy consumption building under different design parameters, and corresponding energy consumption datasets and cost datasets are constructed. Construct an improved support vector machine model based on project priorities; Based on the constructed energy consumption dataset and cost dataset, the improved support vector machine model is trained to obtain the nonlinear relationship between design parameters and energy consumption, as well as the nonlinear relationship between design parameters and cost. Determine an improved multi-objective optimization algorithm based on project priorities; With energy consumption and cost as constraints, the improved multi-objective optimization algorithm is used to perform multi-objective optimization on the nonlinear correlation to obtain the optimal solution.
2. The ultra-low energy consumption building optimization design method according to claim 1, characterized in that, The design parameters of the ultra-low energy consumption building include: heat transfer coefficient of the exterior wall, heat transfer coefficient of the roof, heat transfer coefficient of the partition wall, heat transfer coefficient of the exterior doors and windows, and two or more of the window-to-wall ratio.
3. The ultra-low energy consumption building optimization design method according to claim 1, characterized in that, Each energy consumption dataset contains a set of design parameters and corresponding energy consumption values, where the design parameters include the heat transfer coefficient of the exterior wall, the heat transfer coefficient of the roof, the heat transfer coefficient of the partition wall, the heat transfer coefficient of the exterior doors and windows, and the window-to-wall ratio. Each cost dataset contains a set of design parameters and corresponding costs, where the design parameters include the heat transfer coefficient of the exterior walls, the heat transfer coefficient of the roof, the heat transfer coefficient of the partition walls, the heat transfer coefficient of the exterior doors and windows, and the window-to-wall ratio.
4. The ultra-low energy consumption building optimization design method according to claim 1, characterized in that, The construction of the improved support vector machine model based on project priority includes: Setting Project Priorities: Based on the energy consumption and cost indicators of ultra-low energy building analysis, define project priorities (ICs) accordingly. i,E and IC i,I The value can be: ; ; In the formula, IC i,E For the engineering priority of energy consumption dataset i, IC i,I Let E represent the project priority in cost dataset i, where E represents energy consumption and I represents cost. i Let E be the total energy consumption of energy consumption dataset i. b Let j be the overall energy consumption of the reference building, j be the step size for reducing the energy consumption of the design building relative to the reference building, n be the number of energy consumption datasets i, and C be the total energy consumption of the reference building. i For the cost of cost dataset i, C j Let m be the cost of cost dataset j, and m be the number of cost datasets i. Construct an improved support vector machine model based on project priorities: ; In the formula, w is the weight vector, b is the bias term, ξ is the relaxation factor, and s is the engineering priority vector of dataset i. i Let T be the engineering priority of dataset i, T be the transpose identifier, C be the penalty factor, and ξ be the process priority. i Let y be the relaxation factor for dataset i, st be the constraint condition, and y be the slack factor. i Here, Φ is the labeling function, Φ is the normalization function, and x is the x-axis. i Let i be the design parameters for the i-th group of dataset i.
5. The ultra-low energy consumption building optimization design method according to claim 4, characterized in that, The setting of project priorities includes determining the project priority of the k-th design parameter in dataset i, expressed as: ; In the formula, Let k be the engineering priority of the k-th design parameter in dataset i. Let k be the unit change in the design parameter that affects energy consumption / cost. The impact of a unit change in the j-th design parameter on energy consumption / cost, where n is the number of design parameters, k is the parameter number, and j is the step size by which the energy consumption of the design building decreases relative to the reference building.
6. The ultra-low energy consumption building optimization design method according to claim 1, characterized in that, The method for determining the improved multi-objective optimization algorithm based on project priorities includes: Improve the crossover and / or mutation methods of NSGA-II by adopting engineering priorities based on design parameters; An improved NSGA-II algorithm based on project priority was obtained.
7. The ultra-low energy consumption building optimization design method according to claim 6, characterized in that, The crossover method includes: Determine any i-th set of design parameters x in the dataset i and the j-th group of design parameters x j Conditions for intersection: ; In the formula, s i For the i-th set of design parameters x i Project priority, s j Design parameters x for the j-th group j The project priority, where ε is the project priority threshold; Determine the i-th set of design parameters x i The probability that parameter k is selected for crossover in the design: ; In the formula, For the i-th set of design parameters x i The engineering priority of the k-th design parameter, where n is the number of design parameters. For the i-th set of design parameters x i The engineering priority of the j-th design parameter; For the i-th set of design parameters x i and the j-th group of design parameters x j The k-th design parameter is cross-operated.
8. The ultra-low energy consumption building optimization design method according to claim 6, characterized in that, The mutation methods include: Determine data x i The probability that parameter k is selected for mutation in the design: ; In the formula, For the i-th set of design parameters x i The engineering priority of the k-th design parameter, where n is the number of design parameters, rnd( ) as A random number based on the base number, For the i-th set of design parameters x i The engineering priority of the j-th design parameter, where θ is the design parameter priority threshold.
9. The ultra-low energy consumption building optimization design method according to claim 1, characterized in that, The process of obtaining the optimal solution using the improved multi-objective optimization algorithm, constrained by energy consumption and cost, includes: S1: The N data points generated by the improved model using support vector machines are used as the initial dataset, defined as P0, and the input vector set is... , where x i Let P be the i-th set of design parameters for ultra-low energy consumption buildings. The engineering values are: exterior wall heat transfer coefficient, roof heat transfer coefficient, partition wall heat transfer coefficient, exterior door and window heat transfer coefficient, and window-to-wall ratio for the four orientations. t =P0; S2: For dataset P t Perform a non-dominated sort for each data p. i ∈p t Calculate the number of samples n it dominates i and dominated datasets Where t is the current algebra; S3: Calculate data congestion; S4: Select data based on non-dominated ordering and crowding, and generate offspring data Q through crossover and mutation operations. t ; S5: P t With Q t Merge into a joint dataset R t , for R t Perform a non-dominated sort; S6: Calculate R t Based on data congestion and congestion distance, N data points are selected to form a new dataset P. t+1 ; S7: If the maximum number of generations has been reached, proceed to step S8; otherwise, proceed to step S2. S8: The design parameters corresponding to the optimal output energy consumption and cost.
10. The ultra-low energy consumption building optimization design method according to claim 9, wherein the calculation of data congestion is specifically as follows: ; In the formula, d i For data p i The level of congestion, and For data p i The function value of the adjacent data in the target η, and These are the two extreme values of the target η.
