An artificial intelligence-based lighting device data analysis system and method

Abstract

The application discloses an illumination equipment data analysis system and method based on artificial intelligence, relates to the technical field of data processing and analysis, and comprises the following steps: constructing a lamp initial physical structure, defining lamp bead optical parameters; setting a detection surface position and size, establishing a light intensity spatial distribution model, and defining a light intensity uniformity evaluation function; based on Matlab programming, the detection surface is processed through gridding, the light intensity contribution of a single lamp bead to each grid is calculated, and superposition is performed; a grouping and row strategy is adopted to reduce the calculation complexity, an optimal arrangement scheme is solved through an exhaustive method, and the uniformity indexes before and after optimization are compared; a light source model is established by using Tracepro, light ray tracing parameters are set, light intensity and illumination distribution cloud diagrams before and after optimization are compared, the uniformity improvement effect is quantified, and the algorithm effectiveness is verified. The application can effectively improve the situation that the full exhaustive method is difficult to effectively apply due to excessive calculation in the prior art.

Description

Technical Field

[0001] This invention relates to the field of data processing and analysis technology, specifically to a data analysis system and method for lighting equipment based on artificial intelligence. Background Technology

[0002] LED lighting equipment, with its advantages of high efficiency, energy saving, and long lifespan, is widely used in agriculture, industrial inspection, and other fields. Its lighting effect is closely related to the arrangement of the LED chips; a reasonable arrangement can improve light intensity uniformity and reduce blind spots. Traditional lighting design often relies on empirical formulas or simplified models, which are insufficient to meet the high-precision requirements of complex scenarios. In recent years, with the development of computer simulation technology, simulating light intensity distribution by establishing optical models has become a research hotspot. For example, numerical calculations based on Matlab can quickly process mesh data, while professional optical software such as Tracepro can achieve high-precision ray tracing.

[0003] Existing data analysis methods for lighting equipment have significant limitations when dealing with complex arrangements of multiple LED chips. On one hand, some methods employ exhaustive search to traverse all possible arrangements, resulting in an exponential increase in computational load with the number of LED chips, leading to excessively long optimization times and making them difficult to apply in practical engineering. On the other hand, single-parameter search methods only adjust some key variables, such as spacing or angle, failing to comprehensively consider the coupled effects of multiple factors and easily getting trapped in local optima. Furthermore, traditional methods often treat LED chips as homogeneous light sources, failing to fully consider the differences in optical characteristics of chips with different wavelengths, making it difficult to achieve a balance between spectral distribution and light intensity uniformity in the optimized lighting system. These problems limit the application effectiveness of LED lighting equipment in high-precision scenarios. Summary of the Invention

[0004] The purpose of this invention is to provide a data analysis system and method for lighting equipment based on artificial intelligence, so as to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a data analysis method for lighting equipment based on artificial intelligence, the method comprising the following steps:

[0006] Step 1: Construct the initial physical structure of the lamp, determine the lamp size, the number of rows and columns of LED beads and the reserved space at the edges, and define the optical parameters of the LED beads;

[0007] Step 2: Set the position and size of the detection surface, establish a spatial distribution model of the light intensity of a single LED, and define the light intensity uniformity evaluation function;

[0008] Step 3: Based on Matlab programming, the detection surface is processed by meshing, and the light intensity contribution of a single LED bead to each mesh is calculated and superimposed.

[0009] Step 4: Use a grouping and row-by-row strategy to reduce computational complexity, solve for the optimal layout scheme by exhaustive search, and compare the uniformity index before and after optimization.

[0010] Step 5: Use Tracepro to build a light source model and set ray tracing parameters; compare the light intensity and illuminance distribution cloud maps before and after optimization to quantify the uniformity improvement effect and verify the effectiveness of the algorithm.

[0011] In step 1, the overall dimensions of the lamp are represented as L×W, the horizontal edge allowance is represented as a, the vertical edge allowance is represented as b, and the effective arrangement area is represented as (L-2a)×(W-2b);

[0012] The LED beads are arranged in an M-row N-column array, with the row spacing denoted as d. row =(W-2b) / (M-1), column spacing is represented as d col =(L-2a) / (N-1);

[0013] LED beads are classified by wavelength λ i Classification, with the number of LEDs in each category represented by L. i The corresponding maximum light intensity is represented as I. i,max The half-value angle is represented by θ. i Where i is a positive integer, representing any type of LED bead.

[0014] In step 2, a detection surface with dimensions A×B is set up below the light fixture at a distance h to simulate the light-receiving surface of the crop;

[0015] Calculating the luminous intensity of a single LED: The luminous intensity at a spatial point (x, y, z) satisfies E=I i,max cos n (Φ); where Φ represents the emission angle; n=ln2 / ln(cos(θ) i / 2)) represents the light intensity distribution index;

[0016] The uniformity evaluation function is defined as the ratio of the average light intensity to the maximum light intensity on the detection surface: U=I avg / I max ;

[0017] Among them, I avg Indicates the average light intensity of the detection surface; I max This indicates the maximum light intensity.

[0018] In step 3, the detection surface is divided into P×Q grids, and the center coordinates of each grid are represented as (x...). p ,y q The light intensity value is the light intensity at the center point of the grid.

[0019] For each LED bead (X) j ,Y jZ j ), calculate its position in the grid (x) p ,y q Light intensity contribution at point E: j,pq =I i,max cos n (Φ j,pq ) / r j,pq 2 ;

[0020] Where j∈{1,2,…,M×N}; r j,pq =((x p -X j ) 2 +(y q -Y j ) 2 +h 2 ) 1 / 2 , represents the spatial distance from the j-th LED to the grid (p,q); cos(Φ j,pq )=h / r j,pq , represents the emission angle of the j-th LED from the grid (p,q); j≤M×N;

[0021] Total luminous intensity E pq =Σ j=1 ntotal E j,pq Where ntotal = M × N.

[0022] In step 4, key variables are identified by comparing the uniformity differences of different row / column arrangements;

[0023] When the impact of arranging LED beads in any two rows on the final result is less than a preset threshold:

[0024] The LED beads are divided into K groups according to wavelength, with n beads in each group. k LED beads, distributed in m k Line; where n k This represents the number of LED beads in the k-th group; m k This indicates the row number of the k-th group of LED beads;

[0025] Divide each row of coordinates into S adjacent groups, with T coordinate points in each group. Select the position of the LED bead from each group, and reduce the enumeration quantity from C. N nk Down to C S nk / mk Among them, C N nk C represents the number of all possible ways to select nk positions from N columns of coordinates; S nk / mk This indicates that n is selected from group S after partitioning. k / m kThe number of combinations of groups;

[0026] Exhaustively enumerate the coordinate combinations of each group of LED beads, calculate the corresponding uniformity index U, and select the combination with the largest U as the optimal arrangement.

[0027] Generate LED coordinate distribution maps before and after optimization, and compare the LEDs at each wavelength and the overall uniformity index U. i with U total .

[0028] In step 5, input the coordinates of the LED beads before and after optimization, and define the size of a single LED as s×s×h. led The base model is built according to the actual specifications;

[0029] Set the light intensity distribution curve according to the manufacturer's parameters and distinguish the LED light source attributes according to wavelength type;

[0030] Each LED emits R rays, and the detection surface is located at h below the lamp, with dimensions A×B. The grid accuracy is consistent with Matlab.

[0031] By generating light intensity cloud maps before and after optimization using Tracepro, comparing the light intensity cloud maps before and after optimization, the uniformity improvement effect in the middle area is analyzed, and the edge area is further improved through multi-objective optimization.

[0032] Calculate the average illuminance E of the detection surface avg With the maximum value E max Uniformity U il =E avg / E max Verify the improvement after optimization.

[0033] An artificial intelligence-based lighting equipment data analysis system, comprising a structure definition module, an optical modeling module, a light field calculation module, an optimized layout module, and a simulation verification module;

[0034] The structure definition module is used to construct the initial physical structure of the lamp, determine the lamp size, the number of rows and columns of LED beads, and the reserved space at the edges, and define the optical parameters of the LED beads. The optical modeling module is used to set the position and size of the detection surface, establish a spatial distribution model of the light intensity of a single LED, and define the light intensity uniformity evaluation function. The light field calculation module is used based on Matlab programming to calculate and superimpose the light intensity contribution of a single LED bead to each grid by processing the detection surface through meshing. The optimized arrangement module is used to reduce the computational complexity by adopting a grouping and rowing strategy, solve the optimal arrangement scheme by exhaustive search, and compare the uniformity index before and after optimization. The simulation verification module is used to build a light source model using Tracepro, set ray tracing parameters, compare the light intensity and illuminance distribution cloud maps before and after optimization, quantify the uniformity improvement effect, and verify the effectiveness of the algorithm.

[0035] The structure definition module includes a size definition unit, an array definition unit, and a lamp bead classification unit;

[0036] The size definition unit is used to define the lamp size, edge allowance, and calculate the effective area; the array definition unit is used to set the number of array rows and columns, and calculate the row spacing and column spacing; the lamp bead classification unit is used to classify lamp beads according to wavelength, and define the number of each type, maximum light intensity, and half-value angle.

[0037] The optical modeling module includes a detector surface setting unit, a light intensity model unit, and a uniformity function unit;

[0038] The detection surface setting unit is used to establish a detection surface below the lamp to simulate the light-receiving surface of the crop; the light intensity model unit is used to construct a light intensity model for a single lamp bead; and the uniformity function unit is used to define a uniformity evaluation function.

[0039] The light field calculation module includes a grid division unit, a single lamp contribution unit, and a light intensity superposition unit.

[0040] The grid division unit is used to divide the detection surface into a grid and record the center coordinates; the single lamp contribution unit is used to calculate the light intensity contribution of a single lamp to the grid; the light intensity superposition unit is used to superimpose the contributions of all lamps and calculate the total light intensity of the grid.

[0041] The optimization layout module includes a key variable unit, a grouping dimensionality reduction unit, an exhaustive optimization unit, and a scheme comparison unit.

[0042] The key variable unit is used to identify key row / column arrangement variables that affect uniformity and set grouping thresholds; the grouping dimensionality reduction unit is used to group LEDs by wavelength and coordinates to reduce computational complexity; the exhaustive optimization unit is used to enumerate grouping coordinate combinations, calculate uniformity index values, and select the maximum value scheme; the scheme comparison unit is used to generate coordinate distribution maps before and after optimization and compare them.

[0043] The simulation verification module includes a light source modeling unit, a ray tracing unit, and an effect verification unit.

[0044] The light source modeling unit is used to build a lamp bead model in Tracepro and set the light intensity distribution curve according to the wavelength; the ray tracing unit is used to set the number of rays emitted by each lamp bead and perform ray tracing; the effect verification unit is used to generate light intensity or illuminance cloud map and quantify the improvement of uniformity.

[0045] Compared with existing technologies, the advantages of this invention are as follows: Traditional methods use exhaustive search or single-parameter search, and the computational load increases dramatically with the number of LEDs; this invention reduces the enumeration load from exponential to polynomial level through intelligent grouping and rowing strategies, significantly shortening the optimization time, and is especially suitable for complex array scenarios with multiple LEDs; this invention uses Matlab to quickly calculate the light intensity trend, and then uses Tracepro to simulate the real optical scene. The comparison between the two can quantify the uniformity improvement effect and avoid the problem of "theoretical optimization but actual failure". Attached Figure Description

[0046] Figure 1 This is a schematic diagram illustrating the steps of a lighting equipment data analysis method based on artificial intelligence according to the present invention;

[0047] Figure 2 This is a flowchart illustrating an artificial intelligence-based lighting equipment data analysis system according to the present invention.

[0048] Figure 3 This is a schematic diagram of the initial structure of the lamp according to an embodiment of the present invention;

[0049] Figure 4 This is a schematic diagram comparing the results of LED bead arrangement in different rows according to an embodiment of the present invention;

[0050] Figure 5 This is a partial coordinate diagram of the LED beads before optimization in this embodiment of the invention;

[0051] Figure 6 This is an optimized (partial) LED coordinate diagram according to an embodiment of the present invention;

[0052] Figure 7 This is a model diagram of the light source and receiving surface according to an embodiment of the present invention;

[0053] Figure 8 This is a light intensity distribution diagram of the simulation model before optimization in an embodiment of the present invention;

[0054] Figure 9 This is a light intensity distribution diagram of the optimized simulation model according to an embodiment of the present invention;

[0055] Figure 10 This is an illuminance distribution diagram of the simulation model before optimization in this embodiment of the invention;

[0056] Figure 11 This is an illumination distribution diagram of the optimized simulation model according to an embodiment of the present invention. Detailed Implementation

[0057] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0058] Example: Figures 1-2 As shown, the present invention provides a technical solution, a data analysis method for lighting equipment based on artificial intelligence, the method comprising the following steps:

[0059] Step 1: Construct the initial physical structure of the lamp, determine the lamp size, the number of rows and columns of LED beads and the reserved space at the edges, and define the optical parameters of the LED beads;

[0060] Step 2: Set the position and size of the detection surface, establish a spatial distribution model of the light intensity of a single LED, and define the light intensity uniformity evaluation function;

[0061] Step 3: Based on Matlab programming, the detection surface is processed by meshing, and the light intensity contribution of a single LED bead to each mesh is calculated and superimposed.

[0062] Step 4: Use a grouping and row-by-row strategy to reduce computational complexity, solve for the optimal layout scheme by exhaustive search, and compare the uniformity index before and after optimization.

[0063] Step 5: Use Tracepro to build a light source model and set ray tracing parameters; compare the light intensity and illuminance distribution cloud maps before and after optimization to quantify the uniformity improvement effect and verify the effectiveness of the algorithm.

[0064] In step 1, the overall dimensions of the lamp are represented as L×W, the horizontal edge allowance is represented as a, the vertical edge allowance is represented as b, and the effective arrangement area is represented as (L-2a)×(W-2b);

[0065] The LED beads are arranged in an M-row N-column array, with the row spacing denoted as d. row =(W-2b) / (M-1), column spacing is represented as d col =(L-2a) / (N-1);

[0066] LED beads are classified by wavelength λ i Classification, with the number of LEDs in each category represented by L. i The corresponding maximum light intensity is represented as I. i,max The half-value angle is represented by θ. i Where i is a positive integer, representing any type of LED bead.

[0067] In step 2, a detection surface with dimensions A×B is set up below the light fixture at a distance h to simulate the light-receiving surface of the crop;

[0068] Calculating the luminous intensity of a single LED: The luminous intensity at a spatial point (x, y, z) satisfies E=I i,max cos n (Φ); where Φ represents the emission angle; n=ln2 / ln(cos(θ) i / 2)) represents the light intensity distribution index;

[0069] The uniformity evaluation function is defined as the ratio of the average light intensity to the maximum light intensity on the detection surface: U=I avg / I max ;

[0070] Among them, I avg Indicates the average light intensity of the detection surface; I max This indicates the maximum light intensity.

[0071] In step 3, the detection surface is divided into P×Q grids, and the center coordinates of each grid are represented as (x...). p ,y q The light intensity value is the light intensity at the center point of the grid.

[0072] For each LED bead (X) j ,Y j Z j ), calculate its position in the grid (x) p ,y q Light intensity contribution at point E: j,pq =I i,max cos n (Φ j,pq ) / r j,pq 2 ;

[0073] Where j∈{1,2,…,M×N}; r j,pq =((x p -X j ) 2 +(y q -Y j ) 2 +h 2 ) 1 / 2 , represents the spatial distance from the j-th LED to the grid (p,q); cos(Φ j,pq )=h / r j,pq , represents the emission angle of the j-th LED from the grid (p,q); j≤M×N;

[0074] Total luminous intensity E pq =Σ j=1 ntotal Ej,pq Where ntotal = M × N.

[0075] In step 4, key variables are identified by comparing the uniformity differences of different row / column arrangements;

[0076] When the impact of arranging LED beads in any two rows on the final result is less than a preset threshold:

[0077] The LED beads are divided into K groups according to wavelength, with n beads in each group. k LED beads, distributed in m k Line; where n k This represents the number of LED beads in the k-th group; m k This indicates the row number of the k-th group of LED beads;

[0078] Divide each row of coordinates into S adjacent groups, with T coordinate points in each group. Select the position of the LED bead from each group, and reduce the enumeration quantity from C. N nk Down to C S nk / mk Among them, C N nk C represents the number of all possible ways to select nk positions from N columns of coordinates; S nk / mk This indicates that n is selected from group S after partitioning. k / m k The number of combinations of groups;

[0079] Exhaustively enumerate the coordinate combinations of each group of LED beads, calculate the corresponding uniformity index U, and select the combination with the largest U as the optimal arrangement.

[0080] Generate LED coordinate distribution maps before and after optimization, and compare the LEDs at each wavelength and the overall uniformity index U. i with U total .

[0081] In step 5, input the coordinates of the LED beads before and after optimization, and define the size of a single LED as s×s×h. led The base model is built according to the actual specifications;

[0082] Set the light intensity distribution curve according to the manufacturer's parameters and distinguish the LED light source attributes according to wavelength type;

[0083] Each LED emits R rays, and the detection surface is located at h below the lamp, with dimensions A×B. The grid accuracy is consistent with Matlab.

[0084] By generating light intensity cloud maps before and after optimization using Tracepro, comparing the light intensity cloud maps before and after optimization, the uniformity improvement effect in the middle area is analyzed, and the edge area is further improved through multi-objective optimization.

[0085] Calculate the average illuminance E of the detection surface avg With the maximum value E max Uniformity U il =E avg / E max Verify the improvement after optimization.

[0086] An artificial intelligence-based lighting equipment data analysis system, comprising a structure definition module, an optical modeling module, a light field calculation module, an optimized layout module, and a simulation verification module;

[0087] The structure definition module is used to construct the initial physical structure of the lamp, determine the lamp size, the number of rows and columns of LED beads, and the reserved space at the edges, and define the optical parameters of the LED beads. The optical modeling module is used to set the position and size of the detection surface, establish a spatial distribution model of the light intensity of a single LED, and define the light intensity uniformity evaluation function. The light field calculation module is used based on Matlab programming to calculate and superimpose the light intensity contribution of a single LED bead to each grid by processing the detection surface through meshing. The optimized arrangement module is used to reduce the computational complexity by adopting a grouping and rowing strategy, solve the optimal arrangement scheme by exhaustive search, and compare the uniformity index before and after optimization. The simulation verification module is used to build a light source model using Tracepro, set ray tracing parameters, compare the light intensity and illuminance distribution cloud maps before and after optimization, quantify the uniformity improvement effect, and verify the effectiveness of the algorithm.

[0088] The structure definition module includes a size definition unit, an array definition unit, and a lamp bead classification unit;

[0089] The size definition unit is used to define the lamp size, edge allowance, and calculate the effective area; the array definition unit is used to set the number of array rows and columns, and calculate the row spacing and column spacing; the lamp bead classification unit is used to classify lamp beads according to wavelength, and define the number of each type, maximum light intensity, and half-value angle.

[0090] The optical modeling module includes a detector surface setting unit, a light intensity model unit, and a uniformity function unit;

[0091] The detection surface setting unit is used to establish a detection surface below the lamp to simulate the light-receiving surface of the crop; the light intensity model unit is used to construct a light intensity model for a single lamp bead; and the uniformity function unit is used to define a uniformity evaluation function.

[0092] The light field calculation module includes a grid division unit, a single lamp contribution unit, and a light intensity superposition unit.

[0093] The grid division unit is used to divide the detection surface into a grid and record the center coordinates; the single lamp contribution unit is used to calculate the light intensity contribution of a single lamp to the grid; the light intensity superposition unit is used to superimpose the contributions of all lamps and calculate the total light intensity of the grid.

[0094] The optimization layout module includes a key variable unit, a grouping dimensionality reduction unit, an exhaustive optimization unit, and a scheme comparison unit.

[0095] The key variable unit is used to identify key row / column arrangement variables that affect uniformity and set grouping thresholds; the grouping dimensionality reduction unit is used to group LEDs by wavelength and coordinates to reduce computational complexity; the exhaustive optimization unit is used to enumerate grouping coordinate combinations, calculate uniformity index values, and select the maximum value scheme; the scheme comparison unit is used to generate coordinate distribution maps before and after optimization and compare them.

[0096] The simulation verification module includes a light source modeling unit, a ray tracing unit, and an effect verification unit.

[0097] The light source modeling unit is used to build a lamp bead model in Tracepro and set the light intensity distribution curve according to the wavelength; the ray tracing unit is used to set the number of rays emitted by each lamp bead and perform ray tracing; the effect verification unit is used to generate light intensity or illuminance cloud map and quantify the improvement of uniformity.

[0098] In this embodiment, based on common plant lighting fixtures on the market, the following is adopted: Figure 3 The initial structure of the lamp is shown, and the arrangement of the LED beads is further optimized based on this.

[0099] Based on the selected initial structure, the basic parameters of the luminaire are specified as follows: the overall dimensions of the luminaire are 1150mm × 73mm, and 870 LED beads are evenly placed on the luminaire, arranged in a 6-row, 145-column configuration. Due to wiring and PCB design requirements, 20.6mm is reserved on each of the left and right sides of the luminaire, and 13.5mm is reserved on each of the top and bottom sides. Considering the desired higher uniformity, the 870 LED beads should be more evenly distributed on the remaining 1108.8mm × 46mm plane of the luminaire. The calculated row spacing of the LED beads is 9.2mm, and the column spacing is 7.7mm.

[0100] For the LED chips on the light panel, LED chips manufactured by Youming Optoelectronics Co., Ltd. were selected as the material, and various optical parameters of the LED chips were obtained from the manufacturer's specifications. Regarding the type and quantity of chips, white LED chips were used as the main component, with other colored chips added to meet the characteristic absorption spectrum of the crops and promote rooting and growth. After fitting the characteristic spectrum of the plants, and considering the total number of chips on the panel, the wavelengths and corresponding numbers of the selected LED chips were finally determined.

[0101] The relevant parameters of the LED chips are summarized below. Note that the half-angle value varies slightly for each chip; in subsequent design optimizations, it will be uniformly taken as 120°.

[0102] LED bead number 1: wavelength 450nm, number of beads 28, maximum luminous intensity 1.230cd, half-value angle 131.6°; LED bead number 2: wavelength white light, number of beads 672, maximum luminous intensity 11.71cd, half-value angle 117.2°; LED bead number 3: wavelength 660nm, number of beads 56, maximum luminous intensity 7.945cd, half-value angle 115.6°; LED bead number 4: wavelength 730nm, number of beads 42, maximum luminous intensity 0.0133cd, half-value angle 124.3°; LED bead number 5: wavelength 395nm, number of beads 42, maximum luminous intensity 0.0417cd, half-value angle 124.3°;

[0103] To calculate the uniformity of light intensity from the luminaire, a detection surface needs to be placed below the luminaire to simulate crops receiving light under the luminaire in real-world conditions. Here, the detection surface is placed 200mm below the luminaire, with dimensions of 900mm × 250mm. These parameters will not change during subsequent simulation optimization.

[0104] The light intensity distribution of the lamp on the detection surface is formed by the superposition of the light intensity distribution of individual LED beads (light intensity uniformity design of LED plant tissue culture lamps). The light intensity E emitted by a single LED bead satisfies the following in space: E(x,y,z)=z m I LED / [(xX) 2 +(yY) 2 +z 2 ] (m+2) / 2 In the formula, m = -ln 2 / ln cosθ 1 / 2 cosθ 1 / 2 Let x, y, z be the half-value angle of the LED bead, x, y, z be the coordinates of any point in the luminous intensity distribution area of ​​the LED bead, and X, Y, Z be the position coordinates of the LED bead in space. LED The light intensity at the center of the LED light bead.

[0105] Once the positions of the LED beads are determined, the light intensity at any point in space can theoretically be calculated using the light intensity distribution formula. Given the arrangement of the LED beads on the lamp panel and setting detection points on the detection surface to receive the emitted light intensity, the light intensity distribution on the detection surface can be calculated by superimposing the light intensity values ​​at the detection points.

[0106] Higher light intensity uniformity is beneficial to plant growth. To measure uniformity, the ratio of the average to the maximum light intensity on the probe surface is chosen as a commonly used evaluation function for the uniformity of light intensity received on the probe surface, i.e., U=E. mean / E max In the formula, U represents the uniformity of light intensity distribution on the detection surface, and E... mean E maxThese represent the average light intensity and the maximum light intensity on the detection surface, respectively. Using the previously mentioned light intensity distribution formula, combined with the evaluation formula, the final uniformity calculation result can be obtained.

[0107] Matlab programming approach: Given the spatial distribution of LED light intensity and the maximum light intensity value of the LEDs obtained from the manufacturer, the entire LED lighting fixture can be simulated and modeled.

[0108] Taking a row of monochrome LED beads as an example, firstly, the luminous power of each bead is set to a fixed value. Then, the receiving surface is meshed, and the light intensity at the center of each mesh is defined as the light intensity of that mesh. Finally, the optimal arrangement of the LED beads is determined through a traversal method. The specific steps are as follows:

[0109] (1) Calculate the light intensity of the corresponding receiving surface grid for each LED bead according to the light intensity formula of each LED bead (theoretically, each LED light-emitting particle contributes to the light intensity formation of each grid surface). (2) Assuming there are i LED beads arranged in a row, use an exhaustive method to obtain all possible arrangements of the i LED beads. (3) Superimpose the receiving surface grids of each arrangement in (2) to obtain the corresponding light intensity of the receiving surface grid under each arrangement. (4) Define the light intensity uniformity as the average light intensity of the receiving surface grid divided by the maximum light intensity in the uniform surface grid. Compare the light intensity uniformity in all cases, and the case with the highest light intensity uniformity is the optimal arrangement for that column.

[0110] The arrangement of the 870 LED beads on the overall lighting fixture has been determined by the system's basic parameters; that is, they are arranged in a 6-row, 145-column LED array, and the coordinates of each point in the array are already determined. The description of a particular LED bead's position can be simplified to its row and column number. To simplify the design of the LED bead arrangement, the variables affecting the final uniformity result can be compared individually, and variables with an impact of less than 1% can be ignored.

[0111] Now consider the impact of the LED arrangement in different columns on the final uniformity.

[0112] For the case of two rows of LED beads, randomly select two rows from the six rows on the lamp panel and place 14×2 LED beads of uniform arrangement with identical parameters in each row. Considering the symmetrical relationship between the upper and lower parts of the lamp, a total of 9 different arrangement patterns can be listed. Following the above programming approach, these 9 arrangement patterns are modeled and calculated sequentially in Matlab, and the final uniformity calculation result is obtained, such as... Figure 4As shown in the results, the influence of arranging the LEDs in any two rows on the final result is at most about 0.6%, and the influence on the final uniformity result is almost negligible. Therefore, in the following discussion, the factor of arranging the LEDs in different rows will be ignored, and the LEDs will be placed in different rows only for the convenience of PCB board wiring.

[0113] The specific number of various LED chips has been given. However, due to memory limitations of the enumeration algorithm, it's impossible to directly iterate through and enumerate a single row. Taking 450nm wavelength LED chips as an example, the fitting result shows 28 450nm chips. But selecting 28 positions from 145 coordinate positions in a row... The computational load is clearly too large, exceeding the limits of computer hardware. However, this problem can be solved by dividing the data into rows and groups.

[0114] As discussed above, the arrangement of LEDs in different rows has a negligible impact on the final uniformity. By distributing the 28 LEDs across two rows and dividing the 145 coordinates into groups of five adjacent LEDs, the LED coordinates are selected as if from one of the 29 groups, reducing the computational load to [missing information]. This falls within the computational range of a computer. This approach clearly has both advantages and disadvantages. The disadvantage is that it reduces the selectable coordinates of the LED beads, and the final calculated uniformity result may not be the theoretically optimal value, showing a small discrepancy from the optimal result. The advantage is that it can handle and calculate large amounts of data, and has good universality when dealing with complex types and numbers of LED beads.

[0115] For the lighting fixtures optimized in this modeling, considering both PCB routing issues and the number of LEDs of various wavelengths, the five types of LEDs were grouped as follows: 450nm: 28 LEDs, 2 rows, row numbers 2 and 5; 660nm: 56 LEDs, 4 rows, row numbers 1, 3, 4, and 6; 730nm: 42 LEDs, 3 rows, row numbers 1, 3, and 5; 395nm: 42 LEDs, 3 rows, row numbers 2, 4, and 6.

[0116] The grouping of LED chips was introduced into Matlab, and the arrangement of 870 LED chips was simulated and optimized according to the programming approach described above. The final optimized lamp arrangement and the evaluation results of light intensity uniformity at each wavelength were obtained. The final results were visualized using Matlab's plotting capabilities, resulting in partial LED chip coordinate graphs before and after optimization, as shown below. Figure 5 , Figure 6 .

[0117] Compared with the unoptimized uniform arrangement of LED beads, the results show that the light intensity uniformity of LED beads of various wavelengths is improved by about 5% before and after optimization, indicating that the final optimization results have certain practical significance.

[0118] Comparison of lamp uniformity before and after optimization:

[0119] 450nm: Uniform arrangement 0.8282, optimized arrangement 0.8808;

[0120] (White light): Uniform arrangement 0.8396, optimized arrangement 0.8265;

[0121] 660nm: Uniform arrangement 0.8303, optimized arrangement 0.8817;

[0122] 730nm: Uniform arrangement 0.8300, optimized arrangement 0.8810;

[0123] 395nm: Uniform arrangement 0.8296, optimized arrangement 0.8809;

[0124] Overall: Uniform arrangement 0.8391, Optimized arrangement 0.8297;

[0125] Tracepro Array Simulation Design: After obtaining the coordinate diagrams before and after optimization, to simulate the improvement in light intensity uniformity under real-world conditions, the optical simulation software Tracepro was used to construct a light source model. First, the coordinates of the LED chips were input into Tracepro for simulation. Based on the LED chip specifications provided by Youming Optoelectronics, a base model was established, and the size of a single LED chip was set to 3mm × 3mm × 0.75mm, which was then installed into each LED base. Five different LED types with varying light intensity distributions were set according to the light intensity distribution curves provided by the manufacturer. Considering the large number of LEDs and the long calculation time during tracking calculations, the number of light rays emitted by each LED was set to 1000. The position of the detection surface is consistent with the Matlab simulation, located 200mm below the lamp, with dimensions of 900mm × 250mm. The overall structure is as follows... Figure 7 As shown.

[0126] Using Tracepro to perform ray tracing on the light source model, the light intensity distribution of the light source is obtained as follows: Figure 8 , Figure 9 . Figure 8-9This is a polar coordinate candela plot, where the unit "cd" (candela) of the color bars is the SI unit of luminous intensity. The polar coordinate grid on the right defines spatial orientation. The colors and values ​​on the bars visually represent the magnitude of the luminous intensity of the light source in a specific direction. The plot shows that MATLAB optimization of the LED bead arrangement significantly improves the light intensity uniformity in the central region of the receiving surface. Although the uniformity in the edge region is limited, future improvements can be made through multi-objective optimization, edge bead density adjustment, or optical element design. This result demonstrates the application value of local uniformity optimization in plant supplemental lighting.

[0127] Furthermore, after ray tracing the light source model, an illuminance distribution map can be obtained on the receiving surface. The results before and after optimization are as follows: Figure 10 , Figure 11 As shown. Figure 10-11 The diagram shows the illuminance distribution on the receiving surface. The unit "lux" in the color bars is the SI unit of illuminance. The color bars visually represent the illuminance values ​​at different locations on the receiving surface. The maximum and average illuminance values ​​on the receiving surface can be observed from the diagram. Calculations show that the uniformity before optimization was 0.7142, and the uniformity after optimization was 0.7336, representing an improvement of approximately 3% in array illumination uniformity. The optimization approach proposed in this invention not only optimizes the light intensity distribution but also improves the illuminance uniformity of the receiving surface.

[0128] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. An artificial intelligence-based lighting device data analysis method, characterized by: The method comprises the following steps: Step 1, constructing a lamp initial physical structure, determining lamp size, LED lamp bead arrangement row and column number and edge reserved space, defining LED lamp bead optical parameters; Step 2, setting a detection surface position and size, establishing a single LED light intensity spatial distribution model, and defining a light intensity uniformity evaluation function; The uniformity evaluation function is defined as the ratio of the average light intensity to the maximum light intensity on the detection surface: U = I avg / I max ; where I avg represents the average light intensity of the detection surface; I max represents the maximum light intensity; Step 3, based on Matlab programming, the detection surface is processed by gridding, the light intensity contribution of a single lamp bead to each grid is calculated and superimposed; Step 4, a grouping and row strategy is used to reduce the calculation complexity, an optimal arrangement scheme is solved by an exhaustive method, and the uniformity indexes before and after optimization are compared; In step 4, by comparing the uniformity differences of different row / column arrangements, the key variables are determined; When the influence of lamp bead arrangement in any two rows on the final result is less than a preset threshold value: The lamp beads are divided into K groups according to wavelength, and each group has n k The lamp beads are distributed in m k rows; wherein n k represents the number of LED lamp beads in the kth group; m k represents the number of rows in which the LED lamp beads in the kth group are distributed Divide each row of coordinates into S adjacent groups, each group having T coordinate points, select the lamp bead position from each group, and enumerate the quantity from C N nk to C S nk / mk ; wherein C N nk represents the number of all possible schemes of selecting n k positions from N columns of coordinates; C S nk / mk represents the number of combinations of selecting n k / m k groups from S groups after blocking; The coordinates of each group of lamp beads are exhausted, the corresponding uniformity index U is calculated, and the combination with the maximum U is selected as the optimal arrangement; Generate optimized front light bead coordinate distribution map, compare each wavelength lamp bead and overall uniformity index U i With U total ; Step 5, using Tracepro to establish a light source model and setting light tracing parameters; comparing the light intensity and illumination distribution cloud diagrams before and after optimization, quantifying the uniformity improvement effect, and verifying the effectiveness of the algorithm. 2.The AI-based lighting device data analysis method of claim 1, wherein: In step 1, the overall size of the lamp is represented as LxW, the horizontal edge reservation is represented as a, the vertical edge reservation is represented as b, and the effective arrangement area is represented as (L-2a)x(W-2b); The LED lamp beads are arranged in an M-row N-column array, the row spacing is represented as d row = (W-2b) / (M-1), the column spacing is represented as d col = (L-2a) / (N-1); LED lamp beads according to wavelength λ i Classification, the number of each type of lamp bead is represented as L i , corresponding to the maximum light intensity represented as I i,max , half angle represented as θ i ; wherein i is a positive integer, representing any lamp bead type.

3. The method of claim 2, wherein: In step 2, a detection surface is set at a distance of h below the lamp, with a size of AxB, for simulating the light receiving surface of crops; Calculate the light intensity of single lamp bead: the light intensity of spatial point (x, y, z) satisfies E = I i,max ·cos n (Φ); wherein, Φ represents the light emitting angle; n = ln2 / ln(cos(θ i / 2)), which represents the light intensity distribution index.

4. The method of claim 3, wherein: In step 3, the detection surface is divided into P x Q grids, each grid center coordinate is represented as (x p ,y q ), and the light intensity value is the light intensity of the grid center point; For each LED lamp bead (X j ,Y j ,Z j ), calculate its light intensity contribution at the grid (x p ,y q ): E j,pq =I i,max ·cos n (Φ j,pq ) / r j,pq 2 ; wherein, j∈{1, 2, …, M×N}; r j,pq = ((x p - X j ) 2 + (y q - Y j ) 2 + h 2 ) 1 / 2 , represents the spatial distance from the jth lamp bead to the grid (p, q); cos(Φ j,pq )=h / r j,pq , represents the light-emitting angle of the jth lamp bead to the grid (p, q); j≤M×N; Total light intensity E pq =Σ j=1 ntotal E j,pq ; where ntotal = M x N.

5. The method of claim 4, wherein: In step 5, input the optimized front and rear light bead coordinates, define the size of a single LED as s x s x h led , and the base model is established according to the actual specifications; According to the manufacturer's parameters, set the light intensity distribution curve, and distinguish the LED light source properties according to wavelength types; Each LED lamp bead emits R light rays, the detection surface is located at a distance of h below the lamp, with a size of AxB, and the grid accuracy is consistent with Matlab; Through Tracepro, the light intensity cloud diagrams before and after optimization are generated, the light intensity cloud diagrams before and after optimization are compared, the uniformity improvement effect in the middle area is analyzed, and the edge area is further improved through multi-objective optimization; The average value of the probe surface illuminance E is calculated avg The maximum value E max The uniformity U il =E avg / E max The verification optimization after the lifting range.

6. An artificial intelligence-based lighting device data analysis system applied to the artificial intelligence-based lighting device data analysis method of any one of claims 1-5, characterized in that: The system comprises a structure definition module, an optical modeling module, a light field calculation module, an optimal arrangement module and a simulation verification module; The structure definition module is used for constructing a lamp initial physical structure, determining lamp size, LED lamp bead arrangement row and column number and edge reserved space, and defining LED lamp bead optical parameters; the optical modeling module is used for setting a detection surface position and size, establishing a single LED light intensity spatial distribution model, and defining a light intensity uniformity evaluation function; the light field calculation module is used for programming based on Matlab, processing the detection surface by gridding, calculating the light intensity contribution of a single lamp bead to each grid and superimposing; the optimal arrangement module is used for adopting a grouping and row strategy to reduce calculation complexity, solving an optimal arrangement scheme by an exhaustive method, and comparing the uniformity indexes before and after optimization; and the simulation verification module is used for using Tracepro to establish a light source model and setting light tracing parameters; comparing the light intensity and illumination distribution cloud diagrams before and after optimization, quantifying the uniformity improvement effect, and verifying the effectiveness of the algorithm.

7. The artificial intelligence-based lighting device data analysis system of claim 6, wherein: The structure definition module comprises a size definition unit, an array definition unit and a lamp bead classification unit; The size defining unit is used for defining the size of the lamp, edge reservation, and calculating the effective area; the array defining unit is used for setting the number of rows and columns of the array, and calculating the row spacing and column spacing; the lamp bead classification unit is used for classifying the lamp beads according to wavelength, defining the number of each type, maximum light intensity, and half-value angle; The optical modeling module comprises a detection surface setting unit, a light intensity model unit, and a uniformity function unit; The detection surface setting unit is used for establishing a detection surface below the lamp to simulate the light-receiving surface of crops; the light intensity model unit is used for constructing a single lamp bead light intensity model; and the uniformity function unit is used for defining a uniformity evaluation function.

8. The artificial intelligence-based lighting device data analysis system of claim 7, wherein: The light field calculation module comprises a grid division unit, a single lamp contribution unit, and a light intensity superposition unit; The grid division unit is used for dividing the detection surface into grids and recording the center coordinates; the single lamp contribution unit is used for calculating the light intensity contribution of a single lamp bead to the grid; and the light intensity superposition unit is used for superimposing the contributions of all lamp beads to calculate the total light intensity of the grid; The optimization arrangement module comprises a key variable unit, a grouping dimension reduction unit, an exhaustive optimization unit, and a scheme comparison unit; The key variable unit is used for identifying key row / column arrangement variables affecting uniformity and setting a grouping threshold; the grouping dimension reduction unit is used for grouping lamp beads according to wavelength and coordinates to reduce the calculation complexity; the exhaustive optimization unit is used for enumerating the coordinate combinations of the groups, calculating the uniformity index value, and selecting the maximum value scheme; and the scheme comparison unit is used for generating coordinate distribution diagrams before and after optimization and comparing them.

9. The artificial intelligence-based lighting device data analysis system of claim 8, wherein: The simulation verification module comprises a light source modeling unit, a ray tracing unit, and an effect verification unit; The light source modeling unit is used for establishing a lamp bead model in Tracepro and setting a light intensity distribution curve according to wavelength; the ray tracing unit is used for setting the number of emitted rays of each lamp bead and performing ray tracing; and the effect verification unit is used for generating a light intensity or illumination cloud chart to quantify the uniformity improvement.

Images