CN112307537B

CN112307537B - Natural light vision and non-vision effect coupling optimization method of multi-objective optimization algorithm

Info

Publication number: CN112307537B
Application number: CN202011082604.7A
Authority: CN
Inventors: 吴蔚; 陈启宁
Original assignee: Institute Of Architecture Design & Planning Co ltd Nanjing University; Nanjing University
Current assignee: Institute Of Architecture Design & Planning Co ltd Nanjing University; Nanjing University
Priority date: 2020-10-12
Filing date: 2020-10-12
Publication date: 2022-05-20
Anticipated expiration: 2040-10-12
Also published as: CN112307537A

Abstract

The invention provides a natural light vision and non-vision effect coupling optimization method of a multi-objective optimization algorithm. The method is based on a Rhino & grass shopper parameterization platform, takes Radiance and Daysim as a lighting calculation engine, integrates performance simulation plug-ins such as Honeybee and the like and Python programming language, and combines a multi-objective optimization plug-in Octopus to realize the coupling optimization of visual and non-visual lighting. The essence of the method is that visual and non-visual evaluation indexes of lighting are considered at the same time, and iterative calculation is carried out for a plurality of times through an optimization algorithm to obtain a better solution set range which simultaneously meets visual and non-visual optimization targets. And finally, applying a parameterized model, and visually feeding back result data for a designer to screen according to specific visual and non-visual lighting requirements of the project, thereby providing decision support for construction of a healthy light environment.

Description

Natural light vision and non-vision effect coupling optimization method of multi-objective optimization algorithm

Technical Field

The invention relates to a natural light vision and non-vision effect coupling optimization method of a multi-objective optimization algorithm, and belongs to the technical field of building natural lighting performance simulation.

Background

Natural lighting affects not only the visual effect, but also the circadian rhythms and health of human bodies through non-visual effects, including physiological processes such as hormone secretion, blood pressure, kernel body temperature change and the like. These effects may cause sleep disorders, and may induce Seasonal Affective Disorder (SAD), senile dementia, etc., or may interfere with DNA repair to induce cancer. Further research has shown that non-visual effects differ in terms of response mechanism, influencing factors, viewing direction, etc., as compared to visual effects; namely, the design factors such as the orientation, the room type ratio, the window-wall ratio and the like of the same building have different influences on the vision and the non-vision. Therefore, a high-quality natural light environment satisfying both visual and non-visual needs is created, and the visual and non-visual needs are organically related and considered as a whole.

In recent years, international research on the association of both daylight vision and non-visual effects has been progressively developed. In 2017, Maria L and the like propose a daylighting vision and non-vision combined assessment workflow based on human viewpoints, and three new natural light performance models are utilized to carry out 360-degree immersive space assessment. Qi Dai et al developed a four-channel tunable LED light source mixing method that takes into account both visual and non-visual effects of the illumination. However, these studies only aim at specific background and working condition conditions, and do not consider the diversity of application software in the actual design process, and most of the studies are on qualitative analysis, and lack quantitative analysis studies starting from both visual and non-visual quantitative connection. Therefore, the current research results cannot be developed in the design of real lighting.

The multi-objective optimization algorithm is a method for simultaneously carrying out optimization calculation on a plurality of objective functions, and can realize automatic optimization of a plurality of optimization objectives. The multi-objective optimization is applied in the field of buildings, and design parameters can be adjusted according to specific optimization index requirements such as lighting, thermal comfort, energy consumption and the like, so that multiple optimization objectives under constraint conditions can be realized. For example, Carlucci takes indoor glare and thermal comfort indexes as optimization targets, and applies a genetic algorithm to carry out multi-target optimization on a certain Italy building; negendahl and the like optimize the building maintenance structure by taking natural lighting performance and energy consumption as targets; the periwhite ice takes various natural lighting performance evaluation indexes as optimization targets to research multi-target optimization of the multi-storey office building space in the cold region. Therefore, the multi-objective algorithm can better analyze visual and non-visual factors influencing indoor natural light at the same time, and optimization selection is carried out. In the existing research, optimization research aiming at the non-visual effect of lighting is less, and the existing research mainly focuses on simulation research of the non-visual effect of lighting. In the aspect of optimization research of non-visual effects of lighting, the traditional comparative optimization mode is mainly used at present.

Disclosure of Invention

Aiming at the defects of the prior art in the background art, the invention provides a natural light vision and non-vision effect coupling optimization method of a multi-objective optimization algorithm. The design requirement for meeting the target index of natural light in the target building is met; defining preset parameters of various types corresponding to the light transmission device on the target building as various design parameters; based on the light transmission device on the target building, defining preset visual parameters corresponding to the natural light received inside the target building as visual indexes of the natural light, and defining preset non-visual parameters corresponding to the natural light received inside the target building as non-visual indexes of the natural light;

obtaining first quantitative analysis results of various design parameters and various natural light visual indexes and various natural light non-visual indexes through the steps A to C, and obtaining second quantitative analysis results of various indexes and other indexes and parameters based on various design parameters, various natural light visual indexes and various natural light non-visual indexes; then according to each quantitative analysis result, obtaining each corresponding target design parameter of the target building interior satisfying the natural light visual target index and the natural light non-visual target index, wherein each corresponding target design parameter is used for realizing the design requirement of the target building interior satisfying the natural light target index;

step A, obtaining mapping relations between various design parameters and various natural light visual indexes and various natural light non-visual indexes by establishing a parameterized model, and then entering step B;

b, obtaining a combined solution set among all design parameters by combining all the design parameters and the mapping relation between all the natural light visual indexes and all the natural light non-visual indexes through a multi-objective optimization algorithm, and then entering the step C;

and C, analyzing the combined solution among all the design parameters and the relation between all the natural light visual indexes and all the natural light non-visual indexes through a correlation analysis method and a linear regression analysis method to obtain first quantitative analysis results of all the design parameters and all the natural light visual indexes and all the natural light non-visual indexes respectively, and obtaining second quantitative analysis results of all the indexes and other indexes and parameters respectively based on all the design parameters, all the natural light visual indexes and all the natural light non-visual indexes.

As a preferred embodiment of the present invention: the preset parameters of each type corresponding to the light transmission device on the target building comprise windowsill height, window-to-ground ratio, window interval, window width and window height.

As a preferred embodiment of the present invention: the preset visual parameters corresponding to the natural light received inside the target building comprise natural light autonomous parameters and effective natural lighting illumination.

As a preferred embodiment of the present invention: the preset non-visual parameters corresponding to natural light received inside the target building include effective circadian region percentage and circadian frequency.

As a preferred embodiment of the present invention: and B, establishing a parameterized model for each design parameter through an Rhino & Grasshopper platform, and performing lighting simulation for each natural light visual index and each natural light non-visual index through a Honeybe plug-in, so as to obtain the mapping relation between each design parameter and each natural light visual index and each natural light non-visual index.

As a preferred embodiment of the present invention: and the multi-target optimization algorithm in the step B is a HypE Reduction algorithm.

As a preferred embodiment of the present invention: the correlation analysis method used to obtain the first quantitative analysis result in step C is measured by Pearson correlation coefficient.

As a preferred embodiment of the present invention: the linear regression method for obtaining the second quantitative analysis result in the step C obtains the second quantitative analysis result of each design parameter, each natural light visual index and each natural light non-visual index through the steps C1 to C2 for each natural light visual index and each natural light non-visual index;

c1, selecting one of the natural light visual indexes or the natural light non-visual indexes as an index to be calculated, taking each design parameter as an independent variable and taking the index to be calculated as a dependent variable, then performing linear regression analysis to obtain a linear regression model based on each design parameter and the index to be calculated, and then entering step C2;

step C2. is to perform F-test on the linear regression model to obtain a model formula, and the model formula is used to measure the second quantitative analysis result between the index to be calculated and the design parameter.

As a preferred embodiment of the present invention: and D, obtaining the recommendation intervals of the design parameters according to the combined solution set of the design parameters obtained in the step B and through a preset natural light visual index and a preset natural light non-visual index.

Compared with the prior art, the natural light vision and non-visual effect coupling optimization method of the multi-objective optimization algorithm has the following technical effects by adopting the technical scheme: 1) in terms of index system selection, non-visual indexes adopt the effective Circadian rhythm area percentage (CEA) and Circadian rhythm Frequency (CF) indexes proposed by Konis. The CEA index not only has an index of an illumination threshold value, but also has an index of a stimulation frequency and an effective rhythm area percentage based on illumination history consideration. The CF index can reflect the influence of the light source on the alertness, drowsiness and eyestrain of the user to a certain extent; 2) in view of operability, the lighting calculation of the visual and non-visual effects related by the invention is more convenient, and the operability is stronger. The Grasshopper platform can quickly and conveniently adjust the building scheme for the non-visual effect optimization result. The written python calculation script can simplify parameter setting and realize related index calculation more conveniently. 3) From the aspect of optimization effect, optimization analysis within any time period or whole year range can be carried out in the aspect of optimization analysis time selection of non-visual effect. In the aspect of the spatial distribution optimization analysis of the non-visual effect, the spatial distribution of the relevant indexes can be optimized. In the aspect of quantitative analysis of the optimization result, the optimization result can be subjected to quantitative analysis through correlation analysis and linear regression analysis. In the aspect of visualization of the optimization result, the optimization result can be visualized intuitively; 4) from the practical application effect, the personalized definition and the extension can be carried out according to the use requirement.

The method simultaneously optimizes natural light visual indexes and natural light non-visual indexes through a multi-objective optimization algorithm, obtains a better solution set range which simultaneously meets visual and non-visual optimization targets through repeated iterative calculation, and performs quantitative analysis on the obtained better solution set range by using a correlation analysis and linear regression analysis method, so that an accurate quantitative analysis result and an accurate solution set range interval are obtained.

Drawings

FIG. 1 is a plan view of a classroom in Nanjing, a middle school classroom, according to an embodiment of the present invention;

FIG. 2 is a flow chart of a natural light vision and non-visual effect coupled optimization method of a multi-objective optimization algorithm;

FIG. 3 is a diagram of a parameterized model of a classroom in an embodiment of the invention;

FIG. 4 is a Pareto solution set spatial distribution diagram according to an embodiment of the present invention;

FIG. 5 is a graph comparing pre-and post-optimization visual and non-visual lighting levels in an embodiment of the invention;

FIG. 6 is a scatter distribution diagram of different design parameters and optimization targets in an embodiment of the present invention.

Detailed Description

The traditional natural lighting visual and non-visual effect optimization method has the following defects:

1. cannot optimize aiming at a plurality of lighting indexes at the same time

2. Cannot optimize both visual and non-visual lighting simultaneously

3 parameterized simulation is impossible

4. The natural light non-visual effect optimization can not be carried out by combining with CAAD software.

5. Inaccurate optimization result and failure to obtain quantifiable optimization result

6. The optimization takes longer

The reasons for the above disadvantages are as follows:

1. the traditional natural lighting non-visual effect optimization method only aims at a single target index, and when the number of the optimization targets is more than or equal to 2, a plurality of optimization targets cannot be considered simultaneously.

2. The traditional natural lighting non-visual effect optimization method is optimized by manually changing model parameters, and the work flow cannot be associated with a parameterized model.

3. Currently, visual and non-visual optimization tools and procedures that can be associated with CAAD software are lacking.

4. The traditional natural lighting non-visual effect optimization method can only obtain a rough optimization solution set range, and an accurate solution set interval cannot be calculated.

5. The traditional natural lighting non-visual effect optimization method needs to continuously perform enumeration and comparison aiming at different combinations of independent variables, so that the total time consumption is long. In view of the above, there is a need to introduce a method for optimizing the coupling between natural light vision and non-visual effect, which can be associated with CAAD software, so as to provide a convenient and efficient means for optimizing the natural light vision and non-visual effect for building designers.

The invention aims to link the latest research result of the non-visual assessment model in the international world with a multi-objective optimization algorithm, develop a coupling optimization tool for natural light vision and non-visual effect of buildings on a Grasshopper platform, and realize a coupling optimization method for natural light vision and non-visual effect based on the multi-objective optimization algorithm so as to solve the defects of the existing natural lighting optimization method for vision and non-visual effect.

The following description will explain embodiments of the present invention in further detail with reference to the accompanying drawings. The evaluation index and the software platform used in this embodiment are described in detail below.

1) Evaluation indexes are as follows:

the visual indicators adopt natural light autonomous parameters (daylighting autonomy) and effective natural lighting illuminance (Useful daylighting illmeniance). The two indexes are taken as annual dynamic lighting evaluation indexes, can effectively reflect the change of natural light along with time change, and are verified in many researches. In the aspect of the threshold value range, the DA index takes that the time ratio meeting the natural lighting minimum illumination value of 300lux is more than 75% as the threshold value according to the latest WELL v2 and LEED v4.1 standards; the UDI index is not provided with a unified upper limit value and a unified lower limit value internationally, further verification is needed, and the time ratio of 100lux < UDI <2000lux is temporarily selected to be more than 50% as a threshold value.

Non-visual indicators used were the percentage of effective Circadian area (CEA) and Circadian Frequency (CF) indicators proposed by Konis.

a) The annual stimulation frequency (Stim. freq, SF) is added with the consideration of the sensitization history, so that the effectiveness of the illumination stimulation in a certain time window can be quantified, and the non-visual effect generated by indoor natural light can be intuitively evaluated and predicted. The index arranges a plurality of vectors at even increments on each grid point to represent the direction of view. Typically 8 vectors are specified, and the number of vectors can be increased to achieve higher accuracy. The extent to which the annual stimulation frequency SF maintains a healthy circadian cycle is assessed, typically setting a threshold of 71% for a minimum acceptable stim.freq level, meaning that the requirement for effective stimulation is met for at least 5 days over any 7-day period (stim.freq > -5 d/wk).

b) Circadian Frequency (CF) can be used to assess the extent to which daylighting designs meet the Circadian lighting requirements for using daylight in the WELL standard. Circadian frequency is defined herein as the percentage of days in a particular daily analysis cycle when a given visual vector meets or exceeds a given light stimulus threshold in melatonin illumination eml (equivalent metabolic lux). The index may therefore reflect to some extent the effect of the light source on the alertness, drowsiness and eyestrain of the user.

2) Writing a software platform and a program:

the parametric modeling involved in this embodiment uses modeling software Rhino and a visual programming platform Grasshopper. The computing script is written in Python language and is associated with the Grasshopper platform through a GHpython plug-in. The daylighting calculation simulation and optimization use open source building performance simulation plug-in Honeybe and multi-objective optimization plug-in Octopus on the Grasshopper platform. The Honeybee plug-in uses Daysim and Radiance as lighting simulation engines, is internally provided with a plurality of common CIE standard sky models, can define the optical properties of a plurality of materials, and has been verified by relevant research on the accuracy of lighting simulation.

In order to realize the functions of optimizing calculation of visual and non-visual lighting evaluation indexes and simulating data real-time recording, a set of auxiliary calculation program of the Grasshopper platform is written based on GHpython. The program consists of a parameter input module, a lighting simulation module and an optimization calculation module. The parameter input module generates different parameterized models by adjusting input values of design parameters; the lighting simulation module calculates the non-visual illumination conversion coefficient and the sky type in the selected analysis period by reading meteorological data, so as to realize the lighting simulation calculation of vision and non-vision; and the optimization calculation module carries out iterative transportation through a multi-objective algorithm to obtain a better value range of the design parameters meeting the visual and non-visual requirements at the same time. Through the calculation program of optimization simulation, the evaluation indexes of visual and non-visual lighting can be simulated, optimized in coupling and analyzed quantitatively through the program.

The python script of the lighting simulation module is as follows:

next, a specific implementation process of this embodiment is further described, in this embodiment, a real classroom of a middle school and a high school in tokyo city is selected as an object to be used as a target building to perform a simulation experiment, and in a field measurement study, the classroom of the school meets the requirements of "architectural design specification of middle and primary schools" GBJ 50099-2011 and "architectural lighting design standard" GB 50033-2013. The depth of the classroom is 7m, the bay is 9m, the floor height is 3.5m, the window-to-floor ratio is 0.24, and the specific size is shown in figure 1. As used herein, optical property parameters of the chamber material are based on field measurements, as shown in Table 1.

TABLE 1

In consideration of the complexity of simulation calculation and parametric modeling, only the sill height, the window-to-ground ratio and the window spacing (the vertical centerline spacing between two windows) are selected as design parameters in the optimization in the embodiment. The natural light vision index selects natural light autonomous parameters and effective natural lighting illumination; natural light non-visual indicators employ effective circadian region percentages and circadian frequencies. In order to ensure that the space form result obtained by the optimization calculation meets the relevant specifications, a value range needs to be set for the design parameters during simulation. According to the standard, the ratio of the classroom window is more than 0.2, so the window width and height are set between 1.8 m and 2.3 m, and the windowsill height is between 0.9 m and 1 m. Because the length of the end wall of the side window at the front end of the classroom of the middle and primary schools is not less than 1 meter, the value of the window interval is set to be between 3.0 meters and 3.5 meters in consideration of the value of the window width. Considering the error of lighting simulation, the constraint conditions of the final classroom lighting vision and non-vision optimization design parameter values are shown in table 2:

TABLE 2

The average value of all measuring points is taken as an optimized objective function for each index. Wherein the analysis time interval of the natural light non-visual index is 9: 00-13: 00 according to the minimum standard specified in the WELL standard.

As shown in fig. 2, the following steps a to C are combined with a parameter model and a multi-objective optimization algorithm to obtain quantitative analysis results of design parameters of sill height, window-to-ground ratio, window spacing, and four indexes DA, UDI, CEA, and CF, and then an optimal scheme is selected based on the quantitative analysis results.

A, obtaining mapping relations between various design parameters and various natural light visual indexes and various natural light non-visual indexes by establishing a parameterized model;

in order to realize the mapping between the design parameters and the indexes, a parameterized model is needed for performance simulation. A classroom parameterized model as shown in fig. 3 is established on a Rhino & Grasshopper platform, and dynamic adjustment of windowsill height, window-to-ground ratio and window spacing is realized. When any design parameter is changed, the Honeybee plug-in calls a Radiance kernel and a Daysim kernel to carry out real-time lighting simulation on the parameterized model, automatically feeds back a simulation result, avoids repeated modeling and realizes correlation of evaluation indexes of space morphological information.

When the Honeybee plug-in performs lighting calculation, the Rhino model needs to be converted into the Honeybee Zone model, and then the indoor material is set. Honeybee may provide a call to the Radiance material library.

B, acquiring a better design parameter combination solution set through multi-objective optimization simulation;

the method comprises the steps of performing optimization calculation by adopting an Octopus plug-in based on a HypE Reduction algorithm, taking three design parameters in the embodiment as design parameters of the algorithm, taking four lighting indexes DA, UDI, CEA and CF as optimization targets, respectively inputting the design parameters and the optimization targets into an Octopus calculation module, operating a multi-target optimization algorithm, driving a lighting simulation program to perform repeated iterative calculation on different design parameter combinations, balancing the optimization targets, and seeking for a better design parameter combination solution set.

The optimization has 3 variables, 4 objective functions participate in calculation, and the specific optimization parameter setting of the Octopus plug-in is shown in Table 3. When the parameters are set, the diversity of the solution set and the required calculation amount need to be comprehensively considered, so that the optimal solution is prevented from being lost, and premature convergence is also prevented from being limited to local optimal. After the optimization calculation is completed, the optimization target solution set given by the algorithm can be screened to further obtain a proper design parameter value.

TABLE 3

The optimization takes 70 hours after 35 generations of iterative computation, and 695 groups of data of non-dominated solution sets are obtained in total. The distribution of the Pareto solutions obtained by calculation in the three-dimensional coordinate axis is shown in fig. 4. The curved surface formed by connecting most Pareto optimal solutions is a Pareto front surface, and the Pareto front surface can be formed in the two-dimensional direction and has good convergence. And automatically recording and sorting the result data of the optimization calculation through a GHpython script, and respectively importing all the data into Excel software and Spss software for further data analysis.

The Pareto solution set of generation 35 was analyzed, and there were 32 non-dominant solutions in total, based on the results from the simulation. And 4 groups of solutions with better optimization indexes are selected as a representative optimization scheme, and the optimization target and the design parameter corresponding to each solution are shown in the table 4.

TABLE 4

The optimization scheme in table 4 is plotted as a line graph as shown in fig. 5, in comparison to visual and non-visual indicators of the original classroom. By comparing the values of the indexes before and after optimization, the visual and non-visual indexes are improved simultaneously through optimization. The 3 rd solution forms a better balance among 4 optimization targets, and compared with the original scheme, the DA in the visual index is improved by 16.69%, the UDI is improved by 1.05%, the SF in the non-visual index is improved by 10.34%, and the CF is improved by 9.34%. In general, the non-visual lighting effect is obviously improved.

And C, in order to quantify the relationship between the non-visual indexes and the visual indexes and between the non-visual indexes and the design parameters, analyzing the optimization result of the combined solution set among the design parameters obtained in the step B by adopting a correlation analysis method and a linear regression analysis method which are commonly used in statistics, and obtaining a better value interval of each design parameter by optimizing the distribution condition of data.

1) Correlation analysis

The data were imported into SPSS software, and correlation analysis was performed on Circadian Frequency (CF) with other optimization targets and design parameters, with the results shown in table 5. Pearson correlation in the table represents the strong and weak conditions of correlation, and Pearson coefficients of 0.8-1.0 are extremely strong correlation; 0.6-0.8 are strongly correlated; 0.4-0.6 are moderately correlated; 0.2 to 0.4 are weakly correlated; 0.0-0.2 is irrelevant. Correlation analysis shows that the CF whole-year mean value is in positive correlation with the window-to-ground ratio, the DA mean value and the SF whole-year mean value and is in negative correlation with the windowsill UDI mean value.

TABLE 5

*p<0.05**p<0.01

2) Linear regression analysis

In the SPSS software, a linear regression model was obtained by performing linear regression analysis and automatic linear modeling using the window-to-ground ratio, the window interval, and the sill height as independent variables and the circadian frequency CF as a dependent variable, and the results are shown in table 6.

TABLE 6

From the above table, model R²A value of 0.640 means that window ground ratio, window spacing, sill height may account for 64.0% change in the annual average of CF. The value of the regression coefficient of the window-to-ground ratio is 24.628(t 7.432, p 0.000)<0.01), meaning that the windowed ratio yields a significant positive influence relationship to the full-year mean of CF; the value of the regression coefficient for sill height is-8.971 (t-6.111, p-0.000)<0.01), meaning that sill height has a significant negative impact on the overall annual average of CF, as does the correlation analysis.

When the model is subjected to an F test, the model is found to pass the F test (F is 27.285, p is 0.000<0.05), and the model formula can be obtained: the year round average of CF 80.849+24.628 is higher than +13.560 window spacing-8.971 window steps. Based on the formula, a certain prediction can be made on the annual average value of CF.

3) Preferred interval analysis of design parameters

And importing the data into Excel, taking the design parameters as horizontal coordinates, taking the value of the optimization target in the optimal interval as vertical coordinates, generating a scatter diagram, and intuitively obtaining the optimal interval of each variable according to the distribution condition of the scatter diagram. As shown in fig. 6, values of the design parameters were selected when the annual stimulation frequency stim.freq > 85%. For the annual stimulation frequency SF and the circadian rhythm frequency CF, the value of the time-to-ground ratio is 0.25-0.3; the height value of the windowsill is 0.9-0.95; when the window interval value is 0.4-0.45, the numerical value of the window interval value and the numerical value of the window interval value are always more than 85%. For DA, when the value of the window-to-ground ratio is 0.25-0.3; the height value of the windowsill is 0.9-0.95; when the window interval value is 0.4-0.45, the DA value is always stabilized to be more than 80%. For UDI, the value of the window-to-ground ratio is 0.25-0.2.7; the height value of the windowsill is 0.95-1.0; when the window interval value is 0.4-0.45, the UDI value is always kept above 87.5%.

According to the analysis result, the optimal interval of the design parameters is as follows: the window-to-ground ratio is 0.25-0.3; the height of the windowsill is 0.9-0.95; the distance between the windows is 0.4-0.45.

The embodiments of the present invention have been described in detail with reference to the drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the gist of the present invention.

Claims

1. The natural light vision and non-vision effect coupling optimization method of the multi-objective optimization algorithm is used for realizing that the interior of a target building meets the design requirement of a natural light target index; the method is characterized in that: defining preset parameters of various types corresponding to the light transmission device on the target building as various design parameters; based on the light transmission device on the target building, defining preset visual parameters corresponding to the natural light received inside the target building as visual indexes of the natural light, and defining preset non-visual parameters corresponding to the natural light received inside the target building as non-visual indexes of the natural light;

step A, obtaining mapping relations among various design parameters, various natural light visual indexes and various natural light non-visual indexes by establishing a parameterized model, and then entering step B;

2. The method for coupled optimization of natural light vision and non-visual effects of multi-objective optimization algorithm of claim 1, wherein the parameters of each type preset for the light transmission device on the target building include sill height, window-to-ground ratio, window spacing, window width and window height.

3. The method of claim 1, wherein the preset visual parameters corresponding to the natural light received from the interior of the target building comprise natural light self-contained parameters and effective natural lighting illumination.

4. The method for coupled optimization of natural light vision and non-visual effects of a multi-objective optimization algorithm of claim 1, wherein the preset non-visual parameters corresponding to natural light received inside the target building include effective circadian region percentage and circadian frequency.

5. The method for coupled optimization of natural light vision and non-visual effect of multi-objective optimization algorithm according to claim 1, wherein the mapping relationship in step a is obtained by first establishing a parameterized model for each design parameter through a Rhino & Grasshopper platform, and then performing lighting simulation for each natural light vision index and each natural light non-visual index through a Honeybee plug-in, thereby obtaining the mapping relationship between each design parameter and each natural light vision index and each natural light non-visual index.

6. The method for coupled optimization of natural light vision and non-visual effect of multi-objective optimization algorithm of claim 1, wherein the multi-objective optimization algorithm in step B is HypE Reduction algorithm.

7. The method for coupled optimization of natural light vision and non-visual effects for multi-objective optimization algorithm of claim 1, wherein the correlation analysis method for obtaining the first quantitative analysis result in step C is measured by Pearson correlation coefficient.

8. The coupled optimization method for natural light vision and non-visual effect of multi-objective optimization algorithm of claim 1, wherein the linear regression method for obtaining the second quantitative analysis result in step C obtains the second quantitative analysis result for each design parameter and each natural light vision index and each natural light non-visual index through steps C1 to C2 for each natural light vision index and each natural light non-visual index;

9. The natural light vision and non-visual effect coupled optimization method of the multi-objective optimization algorithm according to claim 1, wherein the quantitative analysis result in the step C further includes a recommended interval of each design parameter, and the recommended interval of each design parameter is obtained through a preset natural light vision index and a preset natural light non-vision index according to the combined solution set of the design parameters obtained in the step B.