Office building envelope optimization design method based on random multi-objective optimization

By considering human behavior uncertainty in architectural design, using NSGA-II algorithm and EnergyPlus software to perform multi-process energy consumption simulation, optimize the design parameters of building envelope structures, and solve the problem of failure to effectively consider uncertainty in the existing technology. , achieving more accurate energy consumption and comfort prediction and more efficient calculations.

CN119047019APending Publication Date: 2024-11-29NINGBO UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410889719.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-04
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

The existing multi-objective optimization method fails to effectively consider the uncertainty of human behavior and design decisions during the architectural design stage, resulting in the deviation of the building performance prediction from the true value and the optimization results are not satisfactory.

Method used

Using the optimization design method of office building enclosure structure based on random multi-objective optimization, by building office building models and determining human behavior models, multi-process energy consumption simulation is performed using NSGA-II algorithm and EnergyPlus software, and integrating three-parameter probability function to describe residents Optimize the design parameters of building enclosure structure for air conditioning control.

Benefits of technology

Improves the accuracy of energy consumption and comfort prediction, provides feasible, excellent and accurate design solutions, significantly improves computing efficiency, and enables more design solutions to be evaluated in a limited time.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119047019A_ABST
    Figure CN119047019A_ABST
Patent Text Reader

Abstract

The invention discloses an office building envelope optimization design method based on random multi-objective optimization, which is characterized in that a multi-objective optimization problem is established for an EnergyPlus office building energy consumption model, optimization objectives are energy consumption and thermal comfort, variables needing to be optimized are design parameters of a building envelope, and the design parameters of the building envelope are optimized. The constraint condition is a design range of variables needing to be optimized; a human behavior model is determined, the human behavior model comprises a displacement model considering the time of arriving at and leaving the office building and an action model for controlling the air conditioner, and then the human behavior model is integrated into the energy consumption simulation process; an NSGA-II algorithm is used for solving a multi-objective optimization problem, a Pareto optimal solution set is obtained, and random optimization is embodied in that multi-process multi-time energy consumption simulation is carried out on each individual in the solving process; the method has the advantages that the uncertainty of human behaviors is considered, and the human behavior model is integrated into the energy consumption simulation process, so that the energy consumption simulation comprises the real situation that people use an office building, and the optimal building envelope design parameters can be obtained.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the fields of building technology, building energy conservation, and in particular to an office building envelope structure optimization design method based on random multi-objective optimization. Background Art

[0002] Multi-objective optimization (MOO) problems are widely used in the early design stage of buildings because they can provide Pareto optimal solutions under conflicting optimization objectives in advance. Rahman Azari et al. used eQuest 3.65 software and life cycle assessment (LCA) method to evaluate the Figure 1 The energy consumption and environmental impact of a low-rise office building were studied, and artificial neural networks and genetic algorithms were used to optimize the building envelope. Evangelia Despoina Giouri et al. developed a near-zero energy building (ZEBs) strategy for a high-rise office building in Greece by optimizing design parameters through multi-objective optimization, using DesignBuilder and EnergyPlus software, and achieved a 33% reduction in annual energy consumption of the building. Mohammadreza Baghoolizadeh et al. used EnergyPlus and JEPLUS software to perform multi-objective optimization on smart sunshades. The optimization results showed that external shading is more effective than internal shading, which can significantly improve the energy efficiency and living comfort of the building. Xudong Zhang et al. optimized window design variables such as window-to-wall ratio and glass type through sensitivity analysis and NSGA-II algorithm, effectively reducing the energy consumption and thermal comfort problems of office buildings, and providing designers with optimized recommended values. However, at this stage, people lack information or knowledge about the true values ​​of some variables, so uncertainty in multi-objective optimization design is difficult to avoid. None of the above existing multi-objective optimization methods take into account uncertainty, including design decisions and human behavior. Uncertainty in design decisions may be due to the uncontrollable nature of design preferences and the lack of information about force majeure related to the project, such as static calculations, fire regulations, etc. Uncertainty in human behavior refers to changes in the habits and patterns of occupants' activities, energy use, etc. Ignoring uncertainty makes the prediction of building performance deviate from the true value during the design phase, and the optimization results are unsatisfactory. Therefore, it is necessary to study multi-objective optimization methods that consider uncertainty. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide an office building envelope structure optimization design method based on stochastic multi-objective optimization, which takes into account the uncertainty of human behavior and integrates the human behavior model into the energy consumption simulation process, so that the energy consumption simulation includes the actual situation of people using office buildings, thereby obtaining the optimal building envelope structure design parameters.

[0004] The technical solution adopted by the present invention to solve the above technical problems is: an office building envelope structure optimization design method based on stochastic multi-objective optimization, characterized by comprising the following steps:

[0005] Step 1: Build an office building model and obtain the EnergyPlus office building energy consumption model; establish a multi-objective optimization problem for the EnergyPlus office building energy consumption model. The optimization objectives of the multi-objective optimization problem are energy consumption and thermal comfort. The variables to be optimized in the multi-objective optimization problem are the design parameters of the building envelope structure. The constraints of the multi-objective optimization problem are the design ranges of the variables to be optimized.

[0006] Step 2: Determine the human behavior model, which includes a displacement model and a motion model. For the displacement model, the time when residents arrive and leave the office building every day conforms to the normal distribution; for the motion model, only the control of the air conditioner by the resident is considered. Turning on and off the air conditioner is an action affected by continuous environmental factors and is described by a three-parameter probability function; then integrate the human behavior model into the energy consumption simulation process;

[0007] Step 3: Use the NSGA-II algorithm to solve the multi-objective optimization problem and obtain the Pareto optimal solution set. The random optimization is reflected in the multi-process energy consumption simulation N times for each individual in the solution process, and the average value of the N simulated energy consumption and the average value of the N discomfort hours are calculated. These two average values ​​are used as indicators for the NSGA-II algorithm to judge the quality of individuals.

[0008] In the step 1, the office building model is built using DesignBuilder software, and the DesignBuilder software outputs the EnergyPlus office building energy consumption model.

[0009] In step 2, the human behavior model is written using a python script.

[0010] In step 2, the three-parameter probability function is expressed as: Among them, P is the probability of action occurrence, e is a natural constant, x is an environmental variable, that is, a continuous environmental factor, u is a threshold, which is used to indicate the possibility of action occurrence starting from a certain point, L is a scale, which is used to indicate the proportional range of environmental stimuli, k is a shape, which is used to indicate the sensitivity of the action to changes in continuous environmental factors, and Δt is the time step.

[0011] When the air conditioner is turned on, u is 26.0℃, L is 5.0℃, and k is 8.0; when the air conditioner is turned off, u is 24.0℃, L is 10.0℃, and k is 3.0.

[0012] In step 2, the human behavior model is integrated into the energy consumption simulation process through the pyenergyplus library provided by the EnergyPlus software.

[0013] Compared with the prior art, the advantages of the present invention are:

[0014] 1) The method of the present invention takes into account the uncertainty of human behavior. The human behavior model is integrated into the energy consumption simulation process in the form of a python script through the pyenergyplus library provided by the EnergyPlus software, so that the simulation includes the actual situation of people using office buildings, and the prediction of energy consumption and comfort will be more accurate.

[0015] 2) In the process of using the genetic algorithm NSGA-II to solve the multi-objective optimization problem, the method of the present invention performs multiple energy consumption simulations of multiple processes for each individual searched, which greatly reduces the influence of randomness, thereby being able to accurately evaluate the influence of different building envelope design parameter selections and human behavior on energy consumption and indoor environmental performance.

[0016] 3) The method of the present invention uses the NSGA-II algorithm to achieve multi-objective optimization of energy consumption and indoor comfort, and obtains the Pareto optimal solution set, which provides a feasible, excellent and accurate solution for decision-making.

[0017] 4) The method of the present invention uses multi-process parallel simulation technology, which significantly improves the computing efficiency and makes it possible to evaluate more design solutions within a limited time. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a general implementation block diagram of the method of the present invention;

[0019] Figure 2 It is a schematic diagram of the three-parameter probability curve for turning on the air conditioner;

[0020] Figure 3 A schematic diagram of the optimization results obtained by performing N = 100 multi-process energy consumption simulations on each design scheme combination (i.e., individuals in the NSGA-II algorithm) in the experiment;

[0021] Figure 4 for Figure 3 Schematic diagram of design parameter results corresponding to the optimization results shown;

[0022] Figure 5 is the p-value cumulative distribution function (CDF) graph of the number of uncomfortable hours obtained in the experiment;

[0023] Figure 6 It is the p-value cumulative distribution function (CDF) graph of energy consumption obtained in the experiment;

[0024] Figure 7 Schematic diagram of the optimization results obtained using the existing multi-objective optimization method. DETAILED DESCRIPTION

[0025] The present invention is further described in detail below with reference to the accompanying drawings.

[0026] The present invention proposes an office building envelope structure optimization design method based on random multi-objective optimization, and its overall implementation block diagram is as follows: Figure 1 As shown, it includes the following steps:

[0027] Step 1: Build an office building model and obtain the EnergyPlus office building energy consumption model; establish a multi-objective optimization problem for the EnergyPlus office building energy consumption model. The optimization objectives of the multi-objective optimization problem are energy consumption and thermal comfort. The variables to be optimized in the multi-objective optimization problem are the design parameters of the building envelope structure. The constraints of the multi-objective optimization problem are the design ranges of the variables to be optimized.

[0028] During the specific implementation, some design parameters of the building envelope structure can also be selected for optimization, and design parameters that have a greater impact on the optimization target should be selected as much as possible, such as window-to-wall ratio, permeability, thickness of exterior wall insulation materials, solar absorption rate of exterior wall insulation materials, population density, etc., and the design range of each design parameter can refer to the standard.

[0029] In step 1 of this embodiment, the office building model is built using DesignBuilder software, and the DesignBuilder software outputs the EnergyPlus office building energy consumption model.

[0030] Step 2: Determine the human behavior model, which includes a displacement model and a motion model. For the displacement model, the time when residents arrive and leave the office building every day conforms to the normal distribution. For the motion model, only the control of the air conditioner by the residents is considered. Turning the air conditioner on and off is an action affected by continuous environmental factors (such as temperature), which is described by a three-parameter probability function. Then integrate the human behavior model into the energy consumption simulation process.

[0031] Here, random factors of human behavior are considered, including the time when occupants arrive and leave the office building, and the three-parameter probability of turning on and off the air conditioner.

[0032] In step 2 of this embodiment, the human behavior model is written using a python script, and the human behavior model is integrated into the energy consumption simulation process through the pyenergyplus library provided by the EnergyPlus software.

[0033] In step 2 of this embodiment, the three-parameter probability function is expressed as: Among them, P is the probability of action occurrence, e is a natural constant, e=2.71..., x is an environmental variable, that is, a continuous environmental factor, u is a threshold, which is used to indicate the possibility of action occurring from a certain point, such as the possibility of turning on the air conditioner from 26.0℃, L is a scale, which is used to indicate the proportional range of environmental stimuli, k is a shape, which is used to indicate the sensitivity of the action to changes in continuous environmental factors, Δt is the time step, and the values ​​of u, L, and k can be adjusted when turning on and off the air conditioner. For example, when turning on the air conditioner, u takes a value of 26.0℃, L takes a value of 5.0℃, and k takes a value of 8.0; when turning off the air conditioner, u takes a value of 24.0℃, L takes a value of 10.0℃, and k takes a value of 3.0.

[0034] Step 3: Use the NSGA-II algorithm in the platypus library of Python to solve the multi-objective optimization problem and obtain the Pareto optimal solution set, wherein the random optimization is reflected in the multi-process energy consumption simulation N times for each individual in the solution process, and the average value of the N simulated energy consumption and the average value of the N discomfort hours are calculated, and these two average values ​​are used as indicators for the NSGA-II algorithm to judge the quality of the individual; wherein N is greater than or equal to 50, and in this embodiment, N is 100.

[0035] In order to further illustrate the feasibility and effectiveness of the method of the present invention, experiments were conducted on the method of the present invention.

[0036] Table 1 gives the design parameter information of the building envelope structure that needs to be optimized for the multi-objective optimization problem in this experiment.

[0037] Table 1 Design parameter information of building envelope structure to be optimized for multi-objective optimization problem

[0038]

[0039] The simulation uses the climate file of Hangzhou City, Zhejiang Province, from July 1 to September 1. The average time for residents to arrive at the office building every day is 8:30, and the average time for residents to leave the office building every day is 20:30, with a standard deviation of 0.83, which is consistent with the normal distribution. The average air conditioning setting temperature is 26℃, with a standard deviation of 0.137, which is also consistent with the normal distribution. The three-parameter probability curve for turning on the air conditioner is as follows: Figure 2 As shown, the horizontal axis is the indoor temperature, and the vertical axis is the probability of turning on the air conditioner; the air conditioner is turned off when leaving the office building.

[0040] For each design scheme combination (i.e., individual in NSGA-II algorithm), N=100 multi-process energy consumption simulations are performed to obtain the average value of 100 energy consumption and the average value of 100 discomfort hours, which are used as indicators for NSGA-II algorithm to judge the quality of individuals. The population size of NSGA-II algorithm is 35 and the number of generations is 60. The optimization results are as follows: Figure 3 As shown, Figure 3 Cooling Energy refers to energy consumption, Not Comfortable Time refers to uncomfortable hours, black dots are individuals searched by the NSGA-II algorithm in the process of solving multi-objective optimization problems, and red dots are Pareto optimal solution sets. The corresponding design parameter results are shown in Figure 4 shown.

[0041] In order to verify the effectiveness of the method of the present invention, a normal distribution test should be performed on the sampled samples of each individual. Compared with other statistical tests, the Shapiro-Wilk test is more suitable for the sample size of this experiment. The p-value of the Shapiro-Wilk test is used to quantify the evidence that the data set does not conform to the normal distribution assumption. The smaller the p-value, the stronger the evidence that the data deviates from the normal distribution. The significance level used in this experiment is 0.05. If the p-value is less than 0.05, the null hypothesis of normal distribution is rejected.

[0042] After the optimization, all individuals were tested for normal distribution. On a single optimization target, 2100 p-values ​​were obtained. The p-value cumulative distribution function (CDF) of the number of uncomfortable hours is shown in the figure below: Figure 5 As shown in the figure, the p-value cumulative distribution function (CDF) of energy consumption is as follows Figure 6 As shown, Figure 5 and Figure 6 The horizontal axis represents the p value, and the vertical axis represents the frequency less than the corresponding horizontal axis value. Figure 5 and Figure 6 It can be seen that, in both optimization objectives, only 5.00% of the individuals have a sample p value less than 0.05, and the assumption of normal distribution is acceptable. Therefore, the method of the present invention is effective.

[0043] In order to prove that the random multi-objective optimization proposed in this invention is superior to the existing multi-objective optimization method (the optimization results of the existing multi-objective optimization method are shown in Figure 2), the random multi-objective optimization method proposed in this invention is superior to the existing multi-objective optimization method (the optimization results of the existing multi-objective optimization method are shown in Figure 2). Figure 7 As shown in Figure 2), the uniformity of the Pareto optimal solution set in the target space can be evaluated by using the Spacing (SP) indicator, which is calculated as follows:

[0044]

[0045] Among them, S represents the Pareto optimal solution set, N represents the number of solutions in the Pareto optimal solution set S, and d * It represents the shortest distance between a solution in the Pareto optimal solution set S and its nearest solution on a certain target (energy consumption or thermal comfort), d represents the average distance of such distance, y i represents the i-th solution in the Pareto optimal solution set S, S / y i Indicates that there is no solution y in the Pareto optimal solution set S except the i-th solution y i All solutions except ij represents the i-th solution y i The value in the jth dimension, a j Indicates that there is no solution y in the Pareto optimal solution set S except the i-th solution y i The value of any solution y in the jth dimension except . It can be seen from the calculation formula of SP that the smaller the value of SP, the better the uniformity of the distribution of the Pareto optimal solution set in the target space. After calculation, the SP value of the Pareto optimal solution set obtained by the existing multi-objective optimization method is 337.39, while the SP value of the Pareto optimal solution set obtained by the random multi-objective optimization method proposed in the present invention is 280.64. Therefore, for the situation of considering human behavior in the optimization design of office buildings, it is recommended to adopt the random multi-objective optimization method proposed in the present invention.

Claims

1. A method for optimizing the envelope structure of an office building based on stochastic multi-objective optimization, characterized in that The following steps are involved: Step 1: Build an office building model and obtain the EnergyPlus office building energy consumption model; establish a multi-objective optimization problem for the EnergyPlus office building energy consumption model. The optimization objectives of the multi-objective optimization problem are energy consumption and thermal comfort. The variables to be optimized in the multi-objective optimization problem are the design parameters of the building envelope structure. The constraints of the multi-objective optimization problem are the design ranges of the variables to be optimized. Step 2: Determine the human behavior model, which includes a displacement model and a motion model. For the displacement model, the time when residents arrive and leave the office building every day conforms to the normal distribution; for the motion model, only the control of the air conditioner by the resident is considered. Turning on and off the air conditioner is an action affected by continuous environmental factors and is described by a three-parameter probability function; then integrate the human behavior model into the energy consumption simulation process; Step 3: Use the NSGA-II algorithm to solve the multi-objective optimization problem and obtain the Pareto optimal solution set. The random optimization is reflected in the multi-process energy consumption simulation N times for each individual in the solution process, and the average value of the N simulated energy consumption and the average value of the N discomfort hours are calculated. These two average values ​​are used as indicators for the NSGA-II algorithm to judge the quality of individuals.

2. The office building envelope optimization design method based on stochastic multi-objective optimization according to claim 1 is characterized in that In the step 1, the office building model is built using DesignBuilder software, and the DesignBuilder software outputs the EnergyPlus office building energy consumption model.

3. The office building envelope optimization design method based on stochastic multi-objective optimization according to claim 1 is characterized in that In step 2, the human behavior model is written using a python script.

4. The office building envelope optimization design method based on stochastic multi-objective optimization according to claim 1 is characterized in that In step 2, the three-parameter probability function is expressed as: Among them, P is the probability of action occurrence, e is a natural constant, x is an environmental variable, that is, a continuous environmental factor, u is a threshold, which is used to indicate the possibility of action occurrence starting from a certain point, L is a scale, which is used to indicate the proportional range of environmental stimuli, k is a shape, which is used to indicate the sensitivity of the action to changes in continuous environmental factors, and Δt is the time step.

5. The office building envelope optimization design method based on stochastic multi-objective optimization according to claim 4 is characterized in that When the air conditioner is turned on, u is 26.0℃, L is 5.0℃, and k is 8.0; when the air conditioner is turned off, u is 24.0℃, L is 10.0℃, and k is 3.

0.

6. The office building envelope optimization design method based on stochastic multi-objective optimization according to claim 1 is characterized in that In step 2, the human behavior model is integrated into the energy consumption simulation process through the pyenergyplus library provided by the EnergyPlus software.