CN115964928A - Multi-objective optimization-based indoor lighting design method and system - Google Patents
Multi-objective optimization-based indoor lighting design method and system Download PDFInfo
- Publication number
- CN115964928A CN115964928A CN202210554839.4A CN202210554839A CN115964928A CN 115964928 A CN115964928 A CN 115964928A CN 202210554839 A CN202210554839 A CN 202210554839A CN 115964928 A CN115964928 A CN 115964928A
- Authority
- CN
- China
- Prior art keywords
- optimization
- model
- design
- establishing
- constraint
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Non-Portable Lighting Devices Or Systems Thereof (AREA)
Abstract
The invention relates to an indoor lighting design method and system based on multi-objective optimization. In the method, the orthogonal test design method is adopted, so that the accurate optimal parameter design solution set can be obtained without carrying out comprehensive tests. In addition, because the genetic algorithm is selected for optimizing the parameters, the optimal solution set of the multi-objective model can be quickly calculated. Therefore, the optimization framework based on the orthogonal test, the regression prediction model and the genetic algorithm can be simplified, the optimal parameter combination solution set can be quickly screened out, and the design and optimization efficiency is greatly improved.
Description
Technical Field
The invention relates to the field of indoor lighting design, in particular to an indoor lighting design method and system based on multi-objective optimization.
Background
With the development of social economy and lighting technology in recent years, people have higher and higher requirements on high efficiency, humanization, health and safety of lighting effects. Moreover, in some special indoor environments, such as narrow spaces like aircraft cabs, train cabs and submarines, the lighting design becomes more complicated and difficult.
In general, lighting designs in these indoor environments often require consideration from multiple dimensions or planes, but when optimizing different objectives, which may be contradictory, optimization of one objective comes at the cost of degradation of the other objective, and thus it is difficult to arrive at a unique optimal solution. Therefore, in the simulation model of the optimization target influenced by a plurality of variables, the optimal parameter combination solution set is rapidly screened out, and the design and optimization efficiency can be greatly improved.
Disclosure of Invention
The invention aims to provide an indoor lighting design method and system based on multi-objective optimization, which can simply and quickly output an optimal parameter solution set while considering a plurality of design targets in the lighting design process.
In order to achieve the purpose, the invention provides the following scheme:
an indoor lighting design method based on multi-objective optimization comprises the following steps:
establishing an optical simulation model;
setting an optimization target and a constraint condition according to design standards and requirements, and selecting optimization variables and levels, wherein the design standards and the requirements are the design standards and the requirements of the industries to which research objects belong, the levels are the values of the optimization variables, and the number of the optimization targets is multiple;
inputting the optimized variables and the optimized levels into the optical simulation model according to an orthogonal test design method for calculation to obtain an orthogonal test table;
establishing a prediction model according to the orthogonal test table, wherein input parameters of the prediction model are the optimization variables, and output parameters are parameter values of an optimization target and a constraint condition;
establishing a multi-objective optimization model according to the prediction model;
and according to the multi-objective optimization model, a genetic algorithm is selected to solve the multi-objective optimization model to obtain an optimal solution set of the multi-objective optimization model, and the optimal solution set is used as a reference for a designer to carry out illumination design.
The invention also provides an indoor lighting design system based on multi-objective optimization, which comprises the following components:
the simulation model establishing module is used for establishing an optical simulation model;
the system comprises a setting and selecting module, a calculating module and a calculating module, wherein the setting and selecting module is used for setting an optimization target and a constraint condition according to a design standard and a requirement, and selecting an optimization variable and a level, the design standard and the requirement are the design standard and the requirement in the industry to which a research object belongs, the level is the value of the optimization variable, and the number of the optimization targets is multiple;
the orthogonal test table acquisition module is used for inputting the optimized variables and the optimized variables into the optical simulation model according to an orthogonal test design method for calculation to obtain an orthogonal test table;
the prediction model establishing module is used for establishing a prediction model according to the orthogonal test table, wherein the input parameters of the prediction model are the optimization variables, and the output parameters are the parameter values of the optimization target and the constraint conditions;
the multi-objective optimization model establishing module is used for establishing a multi-objective optimization model according to the prediction model;
and the model solving module is used for selecting a genetic algorithm to solve the multi-objective optimization model according to the multi-objective optimization model to obtain an optimal solution set of the multi-objective optimization model, and the optimal solution set is used as a reference for designers to carry out illumination design.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the indoor lighting design method and system based on multi-objective optimization provided by the invention firstly establish an optical simulation model, then carry out experimental data acquisition based on an orthogonal experimental design method, then establish a prediction model according to the acquired experimental data, establish a multi-objective optimization model according to the prediction model, and finally solve the multi-objective optimization model based on a genetic algorithm. In the method, the orthogonal test design method is adopted, so that the accurate optimal parameter design solution set can be obtained without carrying out comprehensive tests. In addition, because the genetic algorithm is selected for optimizing the parameters, the optimal solution set of the multi-objective model can be quickly calculated. Therefore, the optimization framework based on the orthogonal test, the regression prediction model and the genetic algorithm can be simplified, the optimal parameter combination solution set can be quickly screened out, and the design and optimization efficiency is greatly improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
Fig. 1 is a flowchart of an indoor lighting design method based on multi-objective optimization according to embodiment 1 of the present invention;
FIG. 2 is a schematic structural diagram of a multilayer sensor model provided in embodiment 1 of the present invention;
fig. 3 is a schematic diagram of an optimal front edge of Pareto provided in embodiment 1 of the present invention;
fig. 4 is a three-dimensional model diagram of a cockpit provided in embodiment 1 of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention aims to provide an indoor lighting design method and system based on multi-objective optimization.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Example 1
The present embodiment provides an indoor lighting design method based on multi-objective optimization, please refer to fig. 1, which includes:
s1, establishing an optical simulation model.
Optionally, the S1 specifically includes:
s11, establishing a three-dimensional model according to a research object, wherein the research object refers to an indoor space environment to be subjected to illumination design.
And S12, importing the three-dimensional model into optical simulation software for optical simulation to obtain an optical simulation model.
Specifically, optical simulation is carried out by using optical simulation software such as Dialux, tracepro, ANSYS Speos and the like, firstly, a three-dimensional model can be established in modeling software such as Catia, solidworks and the like, then, the three-dimensional model is imported into the optical simulation software, and parameters of the optical simulation software such as material, light source, receiving surface and the like are set according to engineering experience and relevant standards and test data of an actual cabin.
And S2, setting an optimization target and a constraint condition according to design standards and requirements, and selecting an optimization variable and level.
According to relevant standards and design requirements in the industry to which the research object belongs, an optimization target and constraint conditions are set, an optimization variable is selected, and the level of the optimization variable, namely the value of the optimization variable, is determined.
And S3, inputting the optimized variables and the optimized levels into the optical simulation model according to an orthogonal test design method for calculation to obtain an orthogonal test table.
And the orthogonal test table is used as a training sample set of a subsequent machine learning prediction model.
And S4, establishing a prediction model according to the orthogonal test table, wherein input parameters of the prediction model are the optimization variables, and output parameters are parameter values of the optimization target and the constraint condition.
And (3) according to an orthogonal test table, bringing the designed optimization variables and levels into the optical simulation model established in the step (S1) one by one, and calculating to obtain the optimization target and the constraint condition parameter values.
Optionally, the optimization target includes parameters for evaluating the indoor light environment, such as glare, spatial light distribution and the like, the constraint condition includes parameters for evaluating the indoor light environment, which are required in standards, such as glare, average illumination of a working surface, uniformity of illumination and the like, and the optimization variable includes lighting design parameters, such as lamp luminous flux, a lamp light distribution curve, and a reflectivity of an interior material.
Due to the fact that different indoor lamps and different environment arrangements are adopted, prediction models of indoor light environment parameters can be different. And (4) preparing to perform machine learning modeling according to the orthogonal test data obtained in the step (S3), and designing a machine learning prediction model with input parameters as optimization variables and output parameters as optimization targets and constraint condition parameter values. The prediction model includes, but is not limited to, least squares regression, polynomial regression, gaussian process regression, neural network, and other regression models.
And S5, establishing a multi-objective optimization model according to the prediction model.
According to the light environment parameter prediction model in the step S4, a multi-objective optimization model is established as shown in the following formula:
the constraint conditions are as follows:
g i (x)≤0,i＝1,2,…,m
h j (x)＝0,j＝1,2,…,k
wherein g (x) is inequality constraint, h (x) is equality constraint, f n (x) For the nth optimization target, m is the number of inequality constraints, k isThe number of equality constraints, i denotes the ith inequality constraint, and j denotes the jth inequality constraint.
And S6, according to the multi-objective optimization model, a genetic algorithm is selected to solve the multi-objective optimization model to obtain an optimal solution set of the multi-objective optimization model, and the optimal solution set is used as a reference for a designer to carry out illumination design.
The solving algorithm comprises genetic algorithms such as NSGA-II, NSGA-III and the like. Finally, the designer may determine the final design parameters based on different design considerations and styles.
In order to make the above-mentioned aspects of the present invention more clearly understood by those skilled in the art, two examples will now be described in detail.
Example (1): the method is implemented on a CATIA/SPEOS simulation platform by taking a certain type of train passenger room as a research object and adopting a bilaterally symmetrical arrangement mode of seats in a carriage. The setting of the optical simulation parameters may refer to standard or actual engineering experience, and is not described in detail in this embodiment. The train passenger room lighting is divided into a top lamp and an annular lamp, wherein the top lamp mainly provides a main lighting appliance required by a visual function; the annular lamp is used for supplementary lighting, and the light distribution of the whole train passenger room environment can be mainly improved.
According to the national railroad industry standard TB/T2917.2-2019, the working surface of a seat area of a train passenger room needs to meet certain illumination and illumination uniformity requirements. Therefore, this example employs, as constraints, the average illuminance (E, con 1) and the illuminance uniformity (U, con 2) on the seat area work plane (0.6 m in front of the back plane, 0.8m high from the ground) at the corner position. The parameter control levels as shown in table 1 were set according to the functionally different and common flat lamp beam angles of the two lamps and actual simulation data. According to CIE 117, the "worst case" of the conventional illumination with a uniform seat array is defined, that is, for an indoor environment with regular, regular and same-directional facilities, the height of the center of two wall surfaces is 1.2m from the ground (the height of eyes of a sitting observer is usually 1.2 m), and the uncomfortable glare feeling is the strongest. In this example, therefore, the glare value UGR (Opt 1) and the spatial light distribution value Feu (Opt 2) in the field of view range (a rectangular field of view of 85 ° vertical and 100 ° horizontal) at the sitting position (1.2 m high) at the end of the car are selected as optimization targets. And (3) the designed optimization variables and levels are gradually added into the optical simulation model established in the step (S1) according to an orthogonal test design method for calculation, so that an orthogonal test data table can be obtained, and the orthogonal test data table is a training sample set which is subsequently used as a machine learning prediction model as shown in a table 2.
TABLE 1 optimized variable set levels
TABLE 2 orthogonal test data sheet
And (3) designing a prediction model with input parameters as optimization variables and output parameters as optimization targets and constraint condition parameter values according to the orthogonal test data table obtained in the step (2). In the present embodiment, data fitting is performed using Multi-layer perceptron regression (MLPR). The method is developed by a single-layer perceptron, and the network structure of the method comprises an input layer, a hidden layer and an output layer. Wherein the hidden layer may be composed of multiple layers, as shown in fig. 2.
Because the multi-layer perceptron model has one or more hidden layers and is fully connected between layers, the MLP has excellent self-learning and non-linear expression capabilities. Here, relu is chosen as the activation function, relu being a simple non-linear function:
y＝Max(0,x)
all parameters to be solved in the MLP model are connection weights and bias vectors between the respective layers. For a particular problem to be solved, the process of determining these parameters is the process of solving the optimization problem. In this example, the Adam algorithm is employed for parameter optimization.
After the prediction model is built, the absolute error value of the test set can be used for verifying the model precision, the smaller the absolute error is, the better the fitting precision of the prediction model is, in the example, samples No. 29-32 are selected as test samples, and the absolute error results of the test samples are listed in Table 3.
TABLE 3 Absolute error analysis results of test samples
As can be seen from table 3 above, each prediction index of each sample achieves high accuracy, and can completely meet the requirement of subsequent multiple optimization.
According to the multi-objective optimization variables, the constraint conditions and the optimization objective set in the step S2, the obtained mathematical model is as follows:
the constraint conditions are as follows:
and (4) solving the multi-objective optimization model by adopting a non-dominated sorting genetic algorithm (NSGA-II algorithm) in combination with the optical parameter prediction model established in the step (S4). The feasible Pareto optimal solution set satisfying the constraint condition can be obtained through iterative computation as shown in table 3. That is, the solution sets shown in table 3 all meet the constraints, and each solution is a non-dominated solution starting from two optimization objectives. The Pareto optimal solution set provides multiple possibilities for designers, who can select optimal parameters from the optimal solution set according to actual needs to design the lighting.
From the data of the optimization result in fig. 3, when a =1000, b =105 °, C =500lm, and d =100 °, UGR is the minimum of 25.1, and the corresponding Feu value is 2.82, the whole vehicle cabin is dark; when a =8000, b =120 °, C =2000lm and d =110 °, feu is the largest and is 10.35, the corresponding UGR value is 31.8, the whole carriage is bright, the beam angle is large, and glare is severe; the designer may decide on different design emphasis and design styles when making the final solution selection. If the designer wishes to focus more on the glare index, the former may be chosen; whereas the latter may be selected if the designer is more emphasised by the spatial visual brightness. Taking these two designs as an example, table 4 shows the error comparison of the two final designs selected.
TABLE 4 error comparison of simulation values and GPR-NSGA-II optimized values
Example (2): the three-dimensional model of the aircraft cockpit shown in fig. 4 is imported into Tracepro, an optical simulation model is built, and parameters such as material, light source and receiving surface are set. These settings may be made with reference to standard or actual engineering experience and are not described in detail in this example.
With reference to relevant design criteria and requirements, the optimization target is set to be the vertical illuminance (G) at eye position (1.2 m), and the illuminance uniformity (denoted as U, respectively) of the two working surfaces 1 And U 2 ) (ii) a The constraint is the average illumination of the two working surfaces (respectively denoted as I) 1 And I 2 ) In which I 1 The value must be between 200 and 400, I 2 The value must be between 100 and 200; the light flux and the beam angle of the two lighting down lamps are selected as optimization variables. Based on the simulation pre-experimental data and market research, the variable levels were selected as shown in table 5 below to meet the constraint requirements. And (3) the designed optimization variables and levels are gradually added into the optical simulation model established in the step (S1) according to an orthogonal test design method for calculation, so that an orthogonal test data table can be obtained, and the orthogonal test data table is a training sample set which is subsequently used as a machine learning prediction model as shown in a table 6.
TABLE 5 optimized variable set levels
TABLE 6 orthogonal test data sheet
And designing a prediction model with input parameters as optimization variables and output parameters as optimization targets and constraint condition parameter values according to the orthogonal test data table obtained in the step S3. In this example, a Gaussian Process Regression (GPR) is used for data fitting. The experimental data set was: x = [ A ] 1 ,B 1 ,A 2 ,B 2 ]，y＝[I 1 ,I 2 ,G,U 1 ,U 2 ]. The distribution for a function can be described by a gaussian process, which is determined by a mean function m (x) and a covariance function k (x, x'):
f(x)～GP(m(x),k(x,x'))
wherein:
m(x)＝E(f(x))
k(x,x')＝E((f(x)-m(x))(f(x')-m(x')))
for ease of calculation, the mean function is typically made 0. The GPR model assumption includes both noise (regression residual) and gaussian process priors, so the general model of gaussian process regression can be expressed as:
y＝f(x)+ε
wherein ε is white Gaussian noise, and ε = N (0, σ) 2 ). The prior distribution of the predicted value y of the prediction model is therefore:
y＝N(0,k(x,x')+σ 2 I)
the training set of the model is: x = (X, y), test set: x = (X, y). The joint probability distribution of the observed value y and the predicted value f (x) is:
wherein K (X, X) is covariance matrix of training set point, K (X, X) = K (X, X) T Covariance matrix for training set points and testing set pointsK (X, X) is the covariance matrix of the test set points. From this, the posterior distribution of the predicted value f (x) can be calculated as:
wherein, the first and the second end of the pipe are connected with each other,
var (f (x)) is the covariance of the predicted value f (x) corresponding to the test point x, and can be calculated by the following formula:
where I is the identity matrix, ε is independent white Gaussian noise, ε = N (0, σ) 2 )，σ 2 Is the variance of gaussian white noise.
The above calculation formula can be used for obtaining a mean value, which is an estimated value of the model, for the gaussian process regression, and the model can calculate the variance of the estimated value, which is closer to reality. Moreover, the selection of the covariance function has a large influence on the accuracy of the model, and the optimal kernel function can be selected by adopting the criterion that the smaller the error of the cross validation is, the better the error is. The hyperparameters in the kernel function and the prior distribution can be generally calculated by adopting maximum likelihood estimation, and the process is not detailed in the embodiment, so that I can be obtained 1 ,I 2 ,G,U 1 ,U 2 With respect to A 1 ,B 1 ,A 2 ,B 2 The machine learning predictive model of (1).
After the prediction model is built, the absolute error value of the test set can be used to verify the model accuracy, and the smaller the absolute error is, the better the fitting accuracy of the prediction model is, in this embodiment, samples No. 29 to 32 are selected as test samples, and the absolute error results of the test samples are listed in table 7.
TABLE 7 Absolute error analysis results of test specimens
As can be seen from table 7 above, each prediction index of each sample achieves very high precision, and can completely meet the requirement of subsequent multiple optimization.
According to the multi-objective optimization variables, the constraint conditions and the optimization objective set in the step S2, the obtained mathematical model is as follows:
the constraint conditions are as follows:
and (4) solving the multi-objective optimization model by adopting a non-dominated sorting genetic algorithm (NSGA-II algorithm) in combination with the optical parameter prediction model established in the step (S4). The feasible Pareto optimal solution set satisfying the constraint condition can be obtained through iterative computation as shown in table 8. That is, the solution sets shown in table 8 all meet the constraints, and each solution is a non-dominated solution from the three optimization objectives. The Pareto optimal solution set provides multiple possibilities for designers, and the designers can select optimal parameters from the optimal solution set according to actual requirements to design the lighting.
TABLE 8 Pareto optimal solution set
When selecting the final scheme, the designer can determine the final scheme according to different design emphasis and design style. If the designer wishes to focus more on the uniformity of the lighting to ensure that the pilot has better performance, the 32 th and 60 th designs can be selected; while designs 6, 19 and 33 may be selected if the designer is more emphasised with visual comfort. Taking the 19 th, 32 th and 60 th designs as examples, table 9 shows the error comparisons of the several final designs selected.
TABLE 9 comparison of error between simulated values and optimized values for GPR-NSGA-II
The present embodiment is directed to a method for quickly searching an optimal parameter design. Thanks to the orthogonal test design method in step S3, an accurate optimal parameter design solution set can be obtained without performing a comprehensive test using this embodiment. Meanwhile, in the prediction model establishing method in the step S4, the model fitting method is not limited in this embodiment, and different regression methods can be selected according to different linear degrees of data, so as to obtain a more accurate prediction model. ( If the optimization target is only influenced by a single variable, the least square linear model can be fitted; if the model is influenced by a plurality of variables simultaneously, regression models such as polynomial regression, gaussian process regression, neural network and the like can be selected according to the absolute error value of the data fitting result )
Then, in this embodiment, a genetic algorithm is selected in step S6 to perform parameter optimization, so that an optimal solution set of the multi-objective model can be quickly calculated. Finally, the optimization framework based on the orthogonal test, the regression prediction model and the genetic algorithm can be simplified, the optimal parameter combination solution set can be rapidly screened out, and the design and optimization efficiency is greatly improved.
Example 2
The embodiment provides an indoor lighting design system based on multi-objective optimization, including:
the simulation model establishing module M1 is used for establishing an optical simulation model;
the setting and selecting module M2 is used for setting an optimization target and a constraint condition according to design standards and requirements, and selecting optimization variables and levels, wherein the design standards and requirements are the design standards and requirements in the industry to which a research object belongs, the levels are the values of the optimization variables, and the number of the optimization targets is multiple;
an orthogonal test table obtaining module M3, configured to input the optimized variables and the levels into the optical simulation model according to an orthogonal test design method for calculation, so as to obtain an orthogonal test table;
the prediction model establishing module M4 is used for establishing a prediction model according to the orthogonal test table, wherein the input parameters of the prediction model are the optimization variables, and the output parameters are the parameter values of the optimization target and the constraint conditions;
the multi-objective optimization model establishing module M5 is used for establishing a multi-objective optimization model according to the prediction model;
and the model solving module M6 is used for selecting a genetic algorithm to solve the multi-objective optimization model according to the multi-objective optimization model to obtain an optimal solution set of the multi-objective optimization model, and the optimal solution set is used as a reference for a designer to carry out illumination design.
Optionally, the formula of the multi-objective optimization model includes:
the constraint conditions are as follows:
g i (x)≤0,i＝1,2,…,m
h j (x)＝0,j＝1,2,…,k
wherein g (x) is inequality constraint, h (x) is equality constraint, f n (x) For the nth optimization target, m is the number of inequality constraints, and k is the number of equality constraints.
For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the foregoing, the description is not to be taken in a limiting sense.
Claims (8)
1. An indoor lighting design method based on multi-objective optimization is characterized by comprising the following steps:
establishing an optical simulation model;
setting an optimization target and a constraint condition according to design standards and requirements, and selecting optimization variables and levels, wherein the design standards and the requirements are the design standards and the requirements in the industry to which a research object belongs, the levels are the values of the optimization variables, and the number of the optimization targets is multiple;
inputting the optimized variables and the level into the optical simulation model according to an orthogonal test design method for calculation to obtain an orthogonal test table;
establishing a prediction model according to the orthogonal test table, wherein input parameters of the prediction model are the optimization variables, and output parameters are parameter values of an optimization target and constraint conditions;
establishing a multi-objective optimization model according to the prediction model;
and according to the multi-objective optimization model, a genetic algorithm is selected to solve the multi-objective optimization model to obtain an optimal solution set of the multi-objective optimization model, and the optimal solution set is used as a reference for a designer to carry out illumination design.
2. The method according to claim 1, wherein the establishing an optical simulation model specifically comprises:
establishing a three-dimensional model according to a research object;
and importing the three-dimensional model into optical simulation software for optical simulation to obtain an optical simulation model.
3. The method of claim 1, wherein the formula of the multi-objective optimization model comprises:
the constraint conditions are as follows:
g i (x)≤0,i＝1,2,…,m
h j (x)＝0,j＝1,2,…,k
wherein g (x) is inequality constraint, h (x) is equality constraint, f n (x) For the nth optimization target, m is the number of inequality constraints, k is the number of equality constraints, i represents the ith inequality constraint, and j represents the jth inequality constraint.
4. The method of claim 1, wherein the optimization objective is to evaluate parameters of indoor light environment, including glare and spatial light distribution.
5. The method of claim 1, wherein the constraint is a parameter for evaluating an indoor light environment, including glare, a work surface average illuminance and an illuminance uniformity.
6. The method of claim 1, wherein the optimization variables are lighting design parameters including lamp luminous flux, lamp light distribution curve and trim material reflectivity.
7. An indoor lighting design system based on multi-objective optimization, comprising:
the simulation model establishing module is used for establishing an optical simulation model;
the system comprises a setting and selecting module, a calculating module and a calculating module, wherein the setting and selecting module is used for setting an optimization target and a constraint condition according to design standards and requirements, and selecting optimization variables and levels, the design standards and the requirements are the design standards and the requirements in the industry to which a research object belongs, the levels are values of the optimization variables, and the optimization targets are multiple;
the orthogonal test table acquisition module is used for inputting the optimized variables and the levels into the optical simulation model according to an orthogonal test design method for calculation to obtain an orthogonal test table;
the prediction model establishing module is used for establishing a prediction model according to the orthogonal test table, wherein the input parameters of the prediction model are the optimization variables, and the output parameters are the parameter values of the optimization target and the constraint conditions;
the multi-objective optimization model establishing module is used for establishing a multi-objective optimization model according to the prediction model;
and the model solving module is used for selecting a genetic algorithm to solve the multi-objective optimization model according to the multi-objective optimization model to obtain an optimal solution set of the multi-objective optimization model, and the optimal solution set is used as a reference for designers to carry out illumination design.
8. The system of claim 7, wherein the formula of the multi-objective optimization model comprises:
the constraint conditions are as follows:
g i (x)≤0,i＝1,2,…,m
h j (x)＝0,j＝1,2,…,k
wherein g (x) is inequality constraint, h (x) is equality constraint, f n (x) For the nth optimization target, m is the number of inequality constraints, k is the number of equality constraints, i represents the ith inequality constraint, and j represents the jth inequality constraint.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210554839.4A CN115964928A (en) | 2022-05-19 | 2022-05-19 | Multi-objective optimization-based indoor lighting design method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210554839.4A CN115964928A (en) | 2022-05-19 | 2022-05-19 | Multi-objective optimization-based indoor lighting design method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115964928A true CN115964928A (en) | 2023-04-14 |
Family
ID=87358658
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210554839.4A Pending CN115964928A (en) | 2022-05-19 | 2022-05-19 | Multi-objective optimization-based indoor lighting design method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115964928A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116306319A (en) * | 2023-05-17 | 2023-06-23 | 天津大学 | Cultural relic illumination glare quantitative evaluation method and system based on genetic algorithm |
| CN116882158A (en) * | 2023-07-06 | 2023-10-13 | 昆明理工大学 | Engineering design-oriented piston-ring set-cylinder structure collaborative optimization design method |
-
2022
- 2022-05-19 CN CN202210554839.4A patent/CN115964928A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116306319A (en) * | 2023-05-17 | 2023-06-23 | 天津大学 | Cultural relic illumination glare quantitative evaluation method and system based on genetic algorithm |
| CN116306319B (en) * | 2023-05-17 | 2023-07-21 | 天津大学 | Cultural relic illumination glare quantitative evaluation method and system based on genetic algorithm |
| CN116882158A (en) * | 2023-07-06 | 2023-10-13 | 昆明理工大学 | Engineering design-oriented piston-ring set-cylinder structure collaborative optimization design method |
| CN116882158B (en) * | 2023-07-06 | 2024-03-26 | 昆明理工大学 | Engineering design-oriented piston-ring set-cylinder structure collaborative optimization design method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115964928A (en) | Multi-objective optimization-based indoor lighting design method and system | |
| CN106228162B (en) | A kind of quick object identification method of mobile robot based on deep learning | |
| CN115642970A (en) | Self-learning channel modeling method and system | |
| Hafiz | Daylighting, space, and architecture: a literature review | |
| Radziszewski et al. | Machine learning algorithm-based tool and digital framework for substituting daylight simulations in early-stage architectural design evaluation | |
| CN111798141A (en) | Method and device for evaluating technology maturity | |
| Ferreira et al. | Genetic assisted selection of RBF model structures for greenhouse inside air temperature prediction | |
| Fountas et al. | A virus-evolutionary multi-objective intelligent tool path optimization methodology for 5-axis sculptured surface CNC machining | |
| Zhang et al. | Intelligent STEP-NC-compliant setup planning method | |
| US8527245B2 (en) | Electronic data processing system for automatic or semi-automatic design and design method | |
| Talbourdet et al. | A knowledge-aid approach for designing high-performance buildings | |
| Aydın et al. | Optimisation of energy consumption and daylighting using building performance surrogate model | |
| Li et al. | A predictive model for daylight performance based on multimodal generative adversarial networks at the early design stage | |
| CN115206455A (en) | Deep neural network-based rare earth element component content prediction method and system | |
| de Carvalho Miranda et al. | Development of an adaptive genetic algorithm for simulation optimization | |
| Yang et al. | Genetic algorithm-based multiobjective optimization for building design | |
| Zhao et al. | Multi-objective optimisation of a free-form building shape to improve the solar energy utilisation potential using artificial neural networks | |
| Si et al. | A novel lambertian-RBFNN for office light modeling | |
| CN115544825A (en) | Method and system for reconstructing deformation of airborne array antenna | |
| CN114841037B (en) | Cab rigidity modal optimization method based on joint sensitivity analysis | |
| CN116975790B (en) | Composite circuit board heat conductivity coefficient prediction method and related equipment | |
| CN116306319B (en) | Cultural relic illumination glare quantitative evaluation method and system based on genetic algorithm | |
| Syberfeldt et al. | Increased robustness of product sequencing using multi-objective optimization | |
| CN113238375B (en) | Method for generating initial structure of free-form surface imaging system based on deep learning | |
| Jayaprakash et al. | Integration of thermo mechanical strains into optimal tolerance design of mechanical assembly using NSGA II and FE simulations |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |