Computer-implemented method for determining the optimal configuration of a light fixture formed by leds
The computer-implemented method optimizes LED configurations to achieve desired spectral and spatial irradiance patterns without additional optical elements, addressing the complexity and cost issues in existing LED luminaire designs by determining the optimal number, type, position, and intensity of LEDs, resulting in simpler and more versatile luminaire manufacturing.
Patent Information
- Application Number
- PCT/ES2024/070283
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-05-10
- Filing Date
- 2024-05-09
- Publication Date
- 2024-11-14
AI Technical Summary
Current LED luminaires require additional optical elements like filters and lenses to achieve controlled spectral and spatial emission patterns, increasing manufacturing complexity and cost, whereas a method to achieve desired lighting patterns using only commercial LEDs without these elements is needed.
A computer-implemented method that determines the optimal configuration of LEDs, including their number, type, position, and intensity, using a genetic algorithm to minimize deviation from a desired spectral irradiance curve and ensure spatial homogeneity, without requiring additional optical elements.
This method allows for the creation of luminaires that can homogeneously irradiate a given area with a desired spectral pattern using only commercially available LEDs, simplifying manufacturing and reducing costs by eliminating the need for filters and lenses, enabling versatile and cost-effective ad-hoc luminaire designs for various applications.
Smart Images

Figure ES2024070283_14112024_PF_FP_ABST
Abstract
Description
[0001] DESCRIPTION
[0002] Computer-implemented method for determining the optimal configuration of a luminaire consisting of LEDs.
[0003] OBJECT OF THE INVENTION
[0004] The present invention falls within the field of lighting, and in particular lighting using LEDs.
[0005] The object of the present invention is a computer-implemented method that allows a luminaire to be designed with a desired illumination pattern, both spectral and spatial. As will be described in detail below, the luminaire is defined by the number, type, position, and current intensity passing through the LEDs that comprise it.
[0006] BACKGROUND OF THE INVENTION
[0007] LED luminaires are now well-known for their use in a wide variety of lighting contexts. As an example, we can refer to the luminaire described in document ES2507440A1, which is specifically designed as an emergency luminaire for use in tunnels.
[0008] In this context, controlled emission patterns with specific spectral and spatial characteristics are frequently required. To achieve this, the most common solution currently consists of adding additional optics to the luminaires that allow emission control (filters to control the spectral dimension, lenses and diffusers for the spatial aspect). Document W02013105046A1 describes a luminaire equipped with a diffusing lens and a plurality of optical elements to redirect the light emitted by the LEDs.
[0009] However, the need to use additional optical elements such as lenses and the like has the disadvantage of increasing the complexity of device manufacturing processes and, consequently, increasing costs.
[0010] In view of this, it would be ideal to have a method that would allow for finding a suitable solution for each case using only commercial LEDs without the need for filters. This solution would allow for obtaining a target emission pattern by combining commercial LED devices, simply by adjusting their spatial distribution in the luminaire and the current intensity at which they are powered, without the need for any additional optical elements.
[0011] DESCRIPTION OF THE INVENTION
[0012] The inventors of the present invention have developed a computer-implemented method that allows determining the configuration of a luminaire to homogeneously irradiate a given area with a desired spectral illumination pattern using only LEDs selected from a predefined set of available LEDs. The resulting luminaire allows generating a desired irradiance curve over the area to be irradiated without the need to add additional optical elements to the LEDs themselves, i.e., without the need to use filters or lenses, and, furthermore, using only LEDs contained in a database of available LEDs. This is very advantageous, since it allows manufacturing ad hoc luminaires for different applications in a much simpler and more versatile manner than by applying filters or lenses.
[0013] The method of the invention essentially comprises the following steps:
[0014] 1. Construct an objective function that calculates, based on the geometry of the luminaire and the area to be irradiated, and a luminaire configuration that includes the total number of LEDs in the luminaire as well as, for each of said LEDs, the type, position and intensity of the current that passes through it, a spectral irradiance curve at various points in the area to be irradiated that corresponds to said luminaire configuration.
[0015] Data on the geometry of the luminaire and the area to be irradiated define the shape, size, and relative position of both elements, and are established at the beginning of the search process for the optimal luminaire. The geometry of the luminaire and the area to be irradiated can be defined using any suitable reference system, although a Cartesian reference system is typically employed. In principle, the luminaire and the area to be irradiated can have different shapes, although according to a particularly preferred embodiment of the invention, the luminaire is flat and the area to be irradiated is also flat, both elements being located in parallel planes.
[0016] As mentioned, the luminaire configuration includes the total number of LEDs in the luminaire, as well as, for each of those LEDs, the type, position, and current intensity passing through it. That is, the luminaire can be compared to a surface on which LEDs are located in different positions. A given luminaire configuration includes information regarding the number of LEDs used, the type of LED located in each particular position of the luminaire, and the current intensity passing through each of those LEDs. Naturally, LEDs supplied with equal or similar current intensities can be electrically grouped.
[0017] Optionally, other additional restrictions can also be set, such as one or more of the following:
[0018] - A restriction on the minimum allowable distance between LEDs. This prevents the process from producing LEDs that must be closer together than their physical size allows.
[0019] - A restriction on the maximum number of LEDs in the luminaire. This prevents solutions with excessive LED counts that have little practical application.
[0020] 2. Apply a genetic algorithm to optimize the objective function with the purpose of minimizing a deviation between the calculated spectral irradiance curve and the desired spectral irradiance curve, as well as a spatial inhomogeneity of the calculated irradiance curve at said points of the plane to be irradiated, the luminaire comprising only LEDs from a predetermined LED database.
[0021] To carry out this step, we first assume a set of initial luminaire configurations, which we will call the initial population, and in which we will call each configuration an 'individual'. This initial population can be determined, for example, by randomly choosing, for each of the individuals in the population, a certain number of LEDs of different types, the positions where these LEDs are located, and the current intensities that power these LEDs. Since the geometry of the luminaire and the area to be irradiated are known, since they are fixed from the beginning of the process, it is possible to use the objective function to determine the spectral irradiance curve corresponding to these particular luminaire configurations. Once the spectral irradiance curve is obtained for each individual in the aforementioned population, the objective function provides, for each configuration in the population, the following results:
[0022] The difference between the calculated spectral irradiance curve and the desired spectral irradiance curve at each point in the area to be irradiated, which ensures that the obtained solution sufficiently resembles the desired target.
[0023] The inhomogeneity of the spectral irradiance curve calculated for the previously determined points in the irradiated area. More specifically, the irradiance at each of these points is calculated, and then the highest and lowest values are selected. The inhomogeneity is calculated taking these two values into account.
[0024] The primary objective of the optimization process is to minimize these quantities. In addition to this objective relating to maximum deviation and maximum inhomogeneity, other additional objectives can be applied. For example, an additional objective could be to achieve a minimum value for the desired spectral coverage, defined as the existence of a specific minimum irradiance in each wavelength subrange within a desired spectral range.
[0025] Ultimately, this test verifies whether the spectral irradiance curve obtained for each individual in the population meets the stated objectives at the defined points in the irradiated area. These points are typically arranged uniformly throughout the irradiated area to ensure, as far as possible, that these objectives are also met over the majority of the area.
[0026] If the solution meets the established conditions, it is considered the optimal solution. Otherwise, the process established in the genetic algorithm used continues, applying transformations to the individuals in the population, and iterating continues until an optimal solution is reached.
[0027] In principle, different types of genetic algorithms could be used to find a solution to the problem posed. However, given the type of variables (a mixture of integers and continuous variables), the fact that there are objective variables to minimize (at least two: deviation from the desired curve and maximum spatial inhomogeneity), and that among the constraints there is at least one linear constraint (maximum number of LEDs), a genetic algorithm of the NSGA (Non-dominated Sorting Genetic Algorithm) type is preferably used.
[0028] Finally, it is important to note that, during the application of the genetic algorithm, the LEDs selected to be part of each luminaire in each iteration are exclusively LEDs from a predetermined LED database. This predetermined LED database may correspond to specific commercially available LED models. The database may include the spectral and angular radiation patterns of the LEDs, as well as any other relevant data. This data can be obtained, for example, from the data sheets corresponding to each LED type included in the database. Naturally, this database can be easily expanded by adding new LED models as they become commercially available.
[0029] 3. Obtain the optimal luminaire configuration, which includes the optimal total number of LEDs in the luminaire as well as, for each of these LEDs, the type, position and intensity of current passing through it.
[0030] Once the genetic algorithm converges to an acceptable solution, the result obtained is a luminaire configuration that constitutes the optimal configuration.
[0031] Ultimately, the result of this process is an optimal luminaire configuration that achieves a spectral irradiance curve close to the desired level and also provides sufficient spatial homogeneity in the irradiated area. Naturally, the final step in this procedure would be the manufacture of the optimal luminaire.
[0032] The present invention is also directed to a computer program comprising instructions that, when executed by a computer, cause the computer to perform the method described in the preceding paragraphs. The program may be in the form of source code, object code, intermediate source code, and object code, for example, in partially compiled form, or in any other form suitable for use in carrying out the processes according to the invention.
[0033] The present invention is also directed to a computer-readable storage medium comprising instructions that, when executed by a computer, cause the computer to perform the method described in the preceding paragraphs. The storage medium may be any entity or device capable of supporting the program. For example, the storage medium could include ROM memory, a CD-ROM memory or a semiconductor ROM memory, or a magnetic recording medium such as a floppy disk or a hard disk, an integrated circuit on which the program is implemented, etc. Alternatively, the storage medium could be a transmissible carrier, for example, an electrical or optical signal transportable through electrical or optical cable, by radio, or by any other means.
[0034] BRIEF DESCRIPTION OF THE DRAWINGS
[0035] Fig. 1 shows a schematic perspective of the luminaire and the area to be irradiated according to an example of the present invention.
[0036] Fig. 2 shows a schematic view of the positions that the LEDs can occupy in the luminaire of this example of the invention.
[0037] Fig. 3 shows the spectral irradiance curve corresponding to each of the LEDs in the LED database used in this example of the invention, as well as the shape of the target solar spectrum AM1,5G.
[0038] Fig. 4 shows the angular irradiance curve corresponding to each of the LEDs in the LED database used in this example of the invention.
[0039] Figs. 5a and 5b show a schematic plan view of the position of the control points in the area to be irradiated used when defining objectives 1 (Fig. 5b), 2 (Fig. 5a) and 3 (Fig. 5b).
[0040] Fig. 6 shows a schematic plan view of the position of the LEDs in the optimal luminaire, as well as the position of the area to be irradiated and its control points.
[0041] Figs. 7a and 7b show, respectively, the total irradiance deviation in each region of the irradiated area and the number of invalid spectral intervals in each region of the irradiated area. Fig. 8 shows the calculated irradiance curve (blue) and the desired irradiance curve (red), corresponding to the AM 1.5G solar spectrum. The inset shows the deviation of the proportion of radiation in each spectral interval of the calculated curve with respect to the proportion that exists in the solar spectrum.
[0042] PREFERRED EMBODIMENT OF THE INVENTION
[0043] To illustrate in more detail what the proposed method consists of, it is described below step by step through a particular application example.
[0044] 1. Problem definition, geometry, constraints and database
[0045] It is desired to obtain a luminaire (1) that provides in an area to be irradiated (3) a radiation curve similar to solar irradiance. Hereinafter, the desired irradiance curve is referred to as AM1.5G, which is the name conventionally assigned to the standard spectrum on the Earth's surface.
[0046] The luminaire (1) is contained in a plane separated by 5 cm from a parallel plane in which the area to be irradiated (3) is contained. The luminaire (1) has dimensions of 5 x 5 cm, while the area to be irradiated (3) has dimensions of 1 x 1 cm and is in a central position with respect to said luminaire (1). Fig. 1 shows schematically the geometric data relating to the luminaire (1) and the area to be irradiated (3).
[0047] An additional restriction is set as a minimum distance between LEDs (2) of 0.5 cm. To implement this restriction, a 5x5 mesh is used in the plane of the luminaire (1) that respects this minimum separation, giving rise to 100 discrete positions numbered as shown in Fig. 2. An additional restriction is also set that limits the number of LEDs (2) in the luminaire to a maximum of 40.
[0048] For this example, a database of commercial LEDs (2) currently in stock is used, comprising 19 different LED models (2). Each of these LEDs (2) emits at different intervals, with the spectral and angular radiation pattern emitted by each LED (2) being stored in said database. Figs. 3 and 4 respectively show the spectral and angular emission patterns corresponding to the LEDs (2) in the database. 2. Definition of the objective function
[0049] In accordance with the invention, the objective function is constructed so that its minimization by means of the genetic algorithm allows obtaining the desired luminaire (1). The variables of this function are: a) The number of LEDs (2) used of each of the 19 types stored in the database. It is an integer variable of range [0,40], b) The current intensity that feeds each group of LEDs (2). A group is defined as a set of LEDs (2) of the same type that are also fed by the same intensity. This facilitates the construction of the luminaire (1) in real practice. It is a continuous variable of range [l m¡n , Imáx] (according to the parameters corresponding to each type of LED). c) The position in the mesh of each of the LEDs (2) used. It is an integer variable of range [1 ,100] (according to the distribution shown above).
[0050] A vector of variables is created according to the previous definition, each vector being the exact description of a luminaire (1), which we will call an individual. An example of a vector of variables for the function would be the following:
[0051] LEDs of each type
[0052] This vector is formed as described below:
[0053] The first set of elements defines the number of LEDs (2) of each type that the luminaire (1) has. In this specific case, the initial series of numbers 1 0 0 2 ... 8 indicates that the luminaire has 1 LED of the first type, 0 LEDs of the second and third types, 2 LEDs of the fourth type, etc. The second set of elements (220 351 400.02 611.3 ... 21.4) describes the current intensity associated with each type of LED (2). In this case, intensities of 220 mA are assigned to the set of LEDs of the first type, 351 mA to the set of LEDs of the second type (although in this case there is none), etc.
[0054] The third set of elements (21 30 78 2... 22) describes the spatial distribution of the individual LEDs (2) in the luminaire (1). In this case, the first LED (which would be of the first type) would be placed in position 21, the next two (which would be of the fourth type) in positions 30 and 78, and so on until the total number of LEDs of this particular individual is completed.
[0055] This vector allows the irradiance calculation to be performed as described in the invention, since it already perfectly describes the luminaire (1). The next step is to establish the desired values for the objective function parameters in order to minimize the difference.
[0056] In this specific problem, when designing a solar simulator, the goal is to meet the current standard for classification as A+A+ in terms of spatial uniformity and spectral coherence. In addition, it is desired to have a total irradiance (in mW / cm 2) similar to that of the AM1.5G spectrum. This defines the three objectives: a) OBJ1 : a certain maximum spatial non-uniformity in the area to be irradiated (3) (specifically, the standard requires less than 1% for class A+). b) OBJ2 : a certain average deviation from the total irradiance of the sun (the standard does not have a criterion in this regard, but it will be set to be less than 5%). c) OBJ3 : a certain deviation of the proportion of spectral radiation from that of the ideal radiation spectrum (specifically, the standard requires that the coupling must be between 0.875 and 1.125, where 1 would be equivalent to the same proportion of radiation as in the case of the sun) between the 6 intervals listed in the following table: Wavelength range ntn
[0057] 300 to 470
[0058] 470 to 561 561 to 657 657 to 772 772 to 919 919 to 1,200
[0059] To establish these objectives mathematically, measurement points are defined in the area to be irradiated (3) that allow monitoring compliance with these conditions. To be as close as possible to the standard, 5 cross-shaped points are established for the second objective (OBJ2) and 9 points for the first and third objectives (OBJ1 and OBJ3), as shown in Figs. 5a and 5b. This choice is based on the fact that the second objective is not as demanding as the other two, and it is sufficient that it is minimized more approximately.
[0060] We proceed to define mathematically the three objectives, thus quantifying the deviation from the ideal luminaire (1) sought:
[0061] Objective 1, which will account for spatial non-uniformity, is calculated through the difference between the maximum (Imax) and minimum (Imin) irradiances found between the 9 points described:
[0062] Imax — Imin
[0063] OBJ1 = • 100 [%]
[0064] Imax + Imin
[0065] Target 2 is defined as the mean deviation from total solar irradiance at the 5 selected points. This restriction is implemented using the following formula: 100 [%]
[0066] Objective 3 represents the sum of the number of 'erroneous' intervals (in which the spectral coupling 0.875-1.125 established by the standard is not met) existing in each of the 9 selected points. This restriction is implemented using the following formula: lido 0.875 and 1.125
[0067] To obtain a luminaire (1) that belongs to class A+A+ in spatial uniformity and spectral coupling, 0BJ1 must be less than 1 and 0BJ3 must be 0. The minimization of 0BJ2, on the other hand, ensures that the configuration of the luminaire (1) allows to achieve an average irradiance close to 1 sun, that is, the irradiance of the AM1.5G spectrum.
[0068] 3. Optimization process
[0069] Once the geometry and objective function have been established, the problem can be limited with the following mathematical characteristics:
[0070] - It is Mixed, that is, it contains a mixture of integer and continuous variables.
[0071] - It is multi-objective, that is, it seeks to minimize objective vahos simultaneously, specifically three.
[0072] - It has linear restrictions, for example: we want the total number of LEDs to not exceed 40.
[0073] The specific genetic algorithm used in this example to solve this problem belongs to the NSGA (Non-dominated Sorting Genetic Algorithm) family. These algorithms are based on generating a population of many diverse individuals (vectors of variables like the one described above) and applying two types of transformations to them:
[0074] • Crossover, mixing between individuals in the population.
[0075] • Mutation: random and individual variations of its parameters.
[0076] A "generation" is defined as the population of individuals obtained as a result of each transformation of the previous one. In each iteration, unfavorable individuals must be excluded according to the defined objectives, and those that minimize them must be benefited, emulating the natural evolutionary process of life. Through successive "generations," the individuals in the conserved population will evolve toward configurations that increasingly approach the goal established for these three objectives.
[0077] Starting with an initial population of 10,000 individuals, after 100 generations, a class A+A+ luminaire is obtained that meets the established criteria, its specific values being:
[0078] • 0BJ1 (spatial non-uniformity): 0.836%
[0079] • 0BJ2 (mean deviation from total solar irradiance): 0.680%
[0080] • 0BJ3 (invalid intervals in spectral coupling): 0
[0081] The solution is shown in Fig. 6, where the position of each of the LEDs (2) in the luminaire (1) is graphically shown by means of a color code representing the type of LED (2). In the lower right part of this figure, a bar graph is added representing the current intensities assigned to each type of LEDs, in orange, superimposed on blue bars indicating their maximum possible value.
[0082] Figs. 7a and 7b show respectively by means of a color code the deviation in the total irradiance as a percentage in each region of the area to be irradiated (3) and the number of invalid spectral intervals in each region of the area to be irradiated (3).
[0083] Fig. 8 shows the spectral irradiance at point (0,0) of the area to be irradiated (3), with an additional bar graph on the right side showing the coupling with the desired AM1.5G radiation curve at that point (0,0).
[0084] The last step of the procedure described would be the manufacturing of the luminaire according to the calculated optimal configuration.
Claims
CLAIMS 1. Computer-implemented method to determine the optimal configuration of a luminaire (1) formed by LEDs (2) with the purpose of achieving a desired spectral irradiance curve on an area to be irradiated (3), where the luminaire (1) lacks optical elements additional to the LEDs themselves (2), characterized in that it comprises: - build an objective function that calculates, based on the geometry of the luminaire (1) and the area to be irradiated (3), and a configuration of the luminaire (1) that includes the total number of LEDs (2) of the luminaire (1) as well as, for each of said LEDs (2), the type, position and intensity of current that passes through it, a spectral irradiance curve at various points of the area to be irradiated (3) that corresponds to said configuration of said luminaire (1); - applying a genetic algorithm to optimize the objective function for the purpose of minimizing a deviation between the calculated spectral irradiance curve and the desired spectral irradiance curve, as well as a spatial inhomogeneity of the calculated spectral irradiance curve at said points of the area to be irradiated (3), the luminaire (1) comprising only LEDs (2) from a predetermined LED database (2); and - obtain the optimal configuration of the luminaire (1), which includes the optimal total number of LEDs (2) of the luminaire (1) as well as, for each of said LEDs (2), the type, position and intensity of current that passes through it.
2. Method according to claim 1, wherein the luminaire (1) is flat and the area to be irradiated (3) is also flat, both being located in parallel planes.
3. Method according to any of the preceding claims, further comprising setting a restriction relating to a minimum allowable distance between LEDs. (2).
4. Method according to any of the preceding claims, further comprising establishing a restriction relating to a maximum number of LEDs (2) of the luminaire.
5. Method according to any of the preceding claims, further comprising establishing a restriction relating to the maximum allowable difference between the calculated spectral irradiance curve and the desired spectral irradiance curve.
6. Method according to any of the preceding claims, further comprising establishing a restriction relating to the minimum desired spectral coverage, being defined as the existence of a certain minimum of irradiance in each subinterval of wavelengths within a desired spectral interval 7. Method according to any of the preceding claims, wherein the genetic algorithm used is an NSGA (Non-dominated Sorting Genetic) algorithm. Algorithm).
8. A computer program comprising instructions that, when the program is executed by a computer, cause the computer to carry out the method of any of claims 1-7.
9. A computer-readable storage medium comprising instructions that, when executed by a computer, cause the computer to perform the method of any of claims 1-7.
Citation Information
Patent Citations
Emergency luminaire for tunnel.
ES2507440A1
Led-based direct-view luminaire with uniform lit appearance
WO2013105046A1
Visible light communication system received power optimizing method based on genetic algorithm
CN104539362A
Method, device and equipment for solving multi-objective optimization problem based on teaching and learning algorithm
CN110147888A
LED light source layout method based on genetic algorithm optimization
CN110457805A
