Light-emitting device and method for controlling light-emitting device

Abstract

The application discloses a lamp bead control method and a light-emitting device. The lamp bead control method is applied to the light-emitting device. The light-emitting device comprises a rigid circuit board and an appearance lamp shell with an appearance surface. The appearance surface faces away from the rigid circuit board. The lamp bead control method determines point cloud data of a middle line track of the appearance surface, an optimal normal vector of a plane where the rigid circuit board is located and a two-dimensional local coordinate system based on a three-dimensional model of the light-emitting device. The vertical mixing distance and non-equidistant coordinates of the lamp beads in a single quadrant are determined based on the two-dimensional local coordinate system and the optimal normal vector. The duty cycle of the pulse width modulation signal for driving the lamp beads in the single quadrant is obtained based on the optimal normal vector and the vertical mixing distance. The non-equidistant coordinates and the duty cycle in the single quadrant are globally mirrored to obtain a space coordinate file and a duty cycle lookup table matrix representing the lamp beads in all quadrants.

Description

Technical Field

[0001] This application relates to the field of optical lighting technology, specifically to a method for controlling the placement of LED beads and a light-emitting device. Background Technology

[0002] In the design of irregularly shaped lighting devices, LEDs with three-dimensional twisted free-form surfaces are typically arranged manually based on experience or evenly spaced. Because rigid circuit boards cannot bend with the surface, the light mixing distance between the LEDs at different positions and the outer casing varies greatly. In addition, due to the Lambertian light-emitting characteristics, this easily causes obvious light and dark patches on the light-emitting surface.

[0003] To address this issue, the industry generally relies on expensive secondary optical lenses, light guides, or light-blocking structures for physical light uniformation. This not only significantly increases material costs and assembly time but also limits the freedom of product appearance design. Summary of the Invention

[0004] This application provides a method for controlling the placement of LED beads and a light-emitting device to alleviate the aforementioned technical problems.

[0005] In a first aspect, this application provides a method for controlling the placement of LED chips. This method is applied to a light-emitting device, which includes a rigid circuit board and an external lamp housing with an external surface facing away from the rigid circuit board. The method includes: determining point cloud data of the centerline trajectory of the external surface, the optimal normal vector of the plane where the rigid circuit board is located, and a two-dimensional local coordinate system based on a three-dimensional model of the light-emitting device; determining the vertical mixing distance and the non-equidistant coordinates of the LED chips in a single quadrant based on the two-dimensional local coordinate system and the optimal normal vector; obtaining the duty cycle of the pulse width modulation signal driving the LED chips in a single quadrant based on the optimal normal vector and the vertical mixing distance; and globally mirroring the non-equidistant coordinates and duty cycle in a single quadrant to obtain a spatial coordinate file representing the LED chips in all quadrants and a duty cycle lookup table matrix.

[0006] Secondly, this application also provides a light-emitting device, which is obtained by the above-described lamp bead arrangement method. Attached Figure Description

[0007] The technical solution and other beneficial effects of this application will become apparent from the following detailed description of specific embodiments in conjunction with the accompanying drawings.

[0008] Figure 1 This is a flowchart illustrating the LED chip deployment method provided in an embodiment of this application.

[0009] Figure 2 This is a schematic diagram of the overall structure of the light-emitting device provided in the embodiments of this application.

[0010] Figure 3An exploded view of the light-emitting device provided in the embodiments of this application.

[0011] Figure 4 This is a schematic diagram of the structure of a light-emitting device within a single quadrant provided in an embodiment of this application.

[0012] Figure 5 This is a structural schematic diagram of the external surface provided in an embodiment of this application.

[0013] Figure 6 A diagram showing the positional relationship between the exterior surface and the light-emitting component, provided for embodiments of this application.

[0014] Figure 7 This is a structural schematic diagram of the center line of the exterior surface provided in an embodiment of this application.

[0015] Figure 8 This is a structural schematic diagram of the light reference point on the center line of the exterior surface provided in the embodiment of this application.

[0016] Figure 9 This is a schematic diagram of the structure of the LED coordinates on the rigid circuit board provided in the embodiments of this application.

[0017] Figure 10 This is a schematic diagram of the structure of the LED on a rigid circuit board provided in the embodiments of this application.

[0018] Figure 11 This is a schematic diagram of the overall structure of the rigid circuit board provided in an embodiment of this application.

[0019] Figure 12 A first-view schematic diagram of the center line of the exterior surface provided in an embodiment of this application.

[0020] Figure 13 A second-view schematic diagram of the centerline of the surface provided in an embodiment of this application.

[0021] Figure 14 A third-view schematic diagram of the center line of the exterior surface provided in an embodiment of this application.

[0022] Icon labels: 10. Exterior light housing; 11. Exterior surface; 12. Center line; 13. Lighting reference point; 20. Bracket; 30. Light-emitting component; 31. Coordinate point; 32. Rigid circuit board; 33. LED bead; 34. First light-emitting angle; 35. Second light-emitting angle; 100. Lighting structure. Detailed Implementation

[0023] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0024] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Features thus defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more unless otherwise explicitly specified.

[0025] In the design of ambient lighting products for irregularly shaped speakers, the exterior surface of the lamp housing typically presents a complex three-dimensional twisted freeform surface. However, due to cost constraints, the circuit board supporting the LED array is usually made of a rigid, inflexible material (such as FR4 board). This structural misalignment between the "three-dimensional curved shell" and the "two-dimensional rigid plane" results in drastic nonlinear fluctuations in the light mixing distance of each LED to the exterior surface, posing a significant challenge to optical uniformity design.

[0026] For such irregularly shaped light-emitting surface products, existing technologies generally adopt a conventional approach combining "equally spaced array distribution" with "uniform constant current or constant power drive." This approach attempts to circumvent complex optical calculations through standardized manufacturing processes, but its physical limitations are exposed when dealing with complex curved surfaces, resulting in serious optical defects. When evenly spaced distributions are projected onto a curved surface, they are very likely to cause overlapping of light spots or create dark areas or dead zones.

[0027] Uncompensated uniform drive cannot cope with the inverse squared decay of illuminance caused by fluctuations in mixing distance.

[0028] Actual test data shows that the wall illumination obtained using existing methods exhibits obvious light and dark blemishes, which seriously affects the visual quality of the product.

[0029] Besides poor optical performance, the existing design process also suffers from efficiency bottlenecks. In the multi-quadrant manual layout mode, designers are prone to interference errors at the quadrant boundaries, and the entire adjustment process relies heavily on manual experience, resulting in extremely low design efficiency and making it difficult to meet the needs of rapid iteration in modern consumer electronics products.

[0030] This application aims to overcome the shortcomings of existing irregularly shaped luminous speaker designs, such as poor optical uniformity, difficulty in structural adaptation, and low R&D efficiency, and provides a luminous speaker structure and its automated lighting layout and electrical closed-loop compensation method. This solution achieves high-quality uniform light emission on a three-dimensional distorted surface by deeply integrating mathematical algorithms and hardware design, eliminating the need for expensive secondary optical components.

[0031] First, to address the physical misalignment between the irregularly shaped exterior surface and the internal inflexible rigid circuit board (FR4), a singular value decomposition (SVD) algorithm based on the least squares criterion is introduced. By analyzing the 3D point cloud data, the optimal spatial normal vector of the rigid circuit board is extracted, reducing the complexity of the 3D surface solution to a 2D local coordinate system for lighting trajectory analysis. This step not only determines the optimal physical installation tilt angle of the hardware, minimizing the nonlinear fluctuations in the light mixing distance, but also avoids, from the algorithm's underlying level, the multi-quadrant interference and boundary ghosting errors that are easily generated by traditional manual layout.

[0032] Secondly, to overcome the uniformity bottleneck of lensless illumination, this solution abandons the traditional equidistant arrangement and constant power driving mode, and innovatively utilizes the inherent Lambertian symmetric light emission characteristics of bare LEDs as a constraint. By establishing a nonlinear boundary equation through simultaneous emission half-angle, the optimal non-equidistant coordinates adapted to the undulations of the curved surface are obtained, ensuring that the light spots are connected end to end on the appearance surface without overlapping gaps. On this basis, the solution further integrates the inverse square law of the mixing distance and the Lambertian correction coefficient to derive an adaptive PWM duty cycle compensation matrix, directly converting the optical uniformity requirement into discrete driving signals for the underlying microcontroller.

[0033] Through the aforementioned synergistic architecture of mechatronics, this invention completely eliminates the need for secondary lenses, light guide plates, and light-blocking structures, while controlling the illuminance deviation of the appearance surface within a very small range. At the same time, it significantly improves the R&D efficiency and optical quality of irregularly shaped light-emitting products, and greatly reduces BOM costs and production time.

[0034] like Figure 1 As shown, this embodiment provides a method for controlling the placement of LED chips in a light-emitting device. The method includes the following steps: Step S10: Based on the three-dimensional model of the light-emitting device, determine the point cloud data of the centerline trajectory of the appearance surface, the optimal normal vector of the plane where the rigid circuit board is located, and the two-dimensional local coordinate system.

[0035] Step S20: Determine the vertical mixing distance and the non-equidistant coordinates of the LEDs in a single quadrant based on the two-dimensional local coordinate system and the optimal normal vector.

[0036] Step S30: Based on the optimal normal vector and vertical mixing distance, obtain the duty cycle of the pulse width modulation signal driving the LED in a single quadrant.

[0037] Step S40: Perform a global mirroring of the non-equally spaced coordinates and duty cycles in a single quadrant to obtain a spatial coordinate file and a duty cycle lookup table matrix representing the LEDs in all quadrants.

[0038] Understandably, the LED placement method provided in this embodiment effectively resolves the assembly interference and height difference contradictions between the irregularly shaped surface and the rigid circuit board, which lead to obvious light and dark patches on the light-emitting surface, by establishing a spatial reference for the rigid circuit board using point cloud data and optimal normal vectors. Based on this, through the coordinated calculation of non-equidistant coordinate arrangement and pulse width modulation duty cycle, precise compensation for Lambertian light attenuation and distance differences is achieved solely through algorithms, eliminating the need for expensive secondary lenses and light guide structures in traditional solutions. Simultaneously, by utilizing global mirroring technology, the precise calculation results of a single quadrant are losslessly extended to the entire domain, significantly improving design efficiency and ensuring high symmetry and consistency of optical output, ultimately achieving high-quality uniform light emission with low material costs.

[0039] It should be noted that, as Figures 2 to 10 As shown, the light-emitting device includes a rigid circuit board 32 and an outer lamp housing 10 having an outer surface 11 facing away from the rigid circuit board 32.

[0040] The centerline trajectory of the exterior surface 11 refers to the center line of the preset light-emitting area on the irregularly shaped speaker exterior light housing 10, which is the geometric reference path for the arrangement of the LED beads 33. In this scheme, the centerline trajectory exists in the form of discrete point cloud data, which is used to characterize the spatial orientation of the three-dimensional curved surface.

[0041] Point cloud data is a collection of massive spatial discrete coordinate points 31 obtained by sampling the centerline trajectory in the 3D model. These spatial discrete coordinate points 31 are the original input for subsequent SVD fitting and geometric calculations.

[0042] The optimal normal vector is the feature vector corresponding to the minimum singular value extracted after performing least-squares fitting on the point cloud data using the Singular Value Decomposition (SVD) algorithm. This feature vector is perpendicular to the optimal mounting plane of the rigid circuit board 32 and serves as the reference axis for establishing the local coordinate system and subsequently calculating the vertical mixing distance and incident angle.

[0043] The two-dimensional local coordinate system is a coordinate system established with the optimal normal vector as the new Z-axis reference. By projecting the global three-dimensional point cloud onto this coordinate system, the complex three-dimensional surface optimization problem is reduced in dimensionality and simplified into a nonlinear programming problem in a two-dimensional plane.

[0044] The vertical mixing distance is the vertical distance measured along the optimal normal vector direction in a two-dimensional local coordinate system between the circuit board plane where the i-th LED 33 is located and the appearance surface 11. This vertical distance is the core parameter for calculating the spot size and illuminance attenuation.

[0045] Non-uniformly spaced coordinates are the non-uniformly distributed position coordinates of the LED beads 33 on the circuit board, calculated by a nonlinear optimization algorithm based on the difference in vertical mixing distance at each point in a two-dimensional local coordinate system in order to satisfy the constraint that the light spots are connected end to end.

[0046] The duty cycle lookup table matrix is ​​a set of pulse width modulation (PWM) duty cycles calculated based on the inverse square law, using the vertical mixing distance of each LED 33 and the Lambert correction coefficient. This duty cycle lookup table matrix is ​​directly sent to the microcontroller to drive the LED 33 to form uniform brightness light on the surface 11.

[0047] Global mirroring utilizes the inherent X=0 and / or Y=0 symmetry planes of the exterior light housing 10 to copy and map the non-equidistant coordinates and duty cycles calculated within a single quadrant (e.g., the first quadrant) to other quadrants. This operation ensures the symmetry and consistency of multi-quadrant lighting layouts and greatly improves design efficiency.

[0048] In this embodiment, point cloud data of the centerline trajectory of the appearance surface 11 is extracted based on the three-dimensional model of the light-emitting device. The SVD algorithm is used to perform least squares fitting on the point cloud, extract the normal vector representing the optimal installation posture of the rigid circuit board 32, and establish a two-dimensional local coordinate system. This transforms the structural misalignment contradiction between the three-dimensional curved surface and the two-dimensional rigid board into a controllable mathematical problem.

[0049] Next, within this two-dimensional coordinate system, a Lambertian light spot boundary constraint equation is established based on the vertical mixing distance of each LED bead 33. The non-equidistant coordinates that make adjacent light spots connect end to end are solved through nonlinear iteration, completely avoiding the dark areas or overlaps generated by the traditional equidistant arrangement on the curved surface.

[0050] Subsequently, the incident angle of the emitted light of each LED bead 33 is calculated based on the optimal normal vector to obtain the Lambert correction coefficient. Combined with the inverse square law of vertical mixing distance, the PWM duty cycle of driving LED bead 33 is derived. Optical attenuation is offset by electrical compensation alone under lensless conditions, thus achieving uniform illumination.

[0051] Finally, by utilizing the inherent symmetry of the appearance, the precise calculation results of a single quadrant are globally mirrored, and the spatial coordinate file and duty cycle lookup table matrix of the global LED beads 33 are generated with one click. The results are then simultaneously output to the CAD layout and the microcontroller firmware, thereby achieving a low-cost, uniform wall-washing light effect on the irregularly shaped surface 11 without the need for expensive secondary optical components.

[0052] In some embodiments, determining the point cloud data of the centerline trajectory of the appearance surface 11, the optimal normal vector of the plane where the rigid circuit board 32 is located, and the two-dimensional local coordinate system based on the three-dimensional model of the light-emitting device includes: determining the point cloud data of the centerline trajectory of the appearance surface 11 based on the three-dimensional model of the light-emitting device; obtaining a right singular vector matrix based on the point cloud data; determining the eigenvector corresponding to the smallest singular value in the right singular vector matrix as the optimal normal vector of the mounting plane of the rigid circuit board 32 based on the least squares fitting criterion; and constructing a rotation transformation matrix based on the optimal normal vector to transform the point cloud data into a two-dimensional local coordinate system based on the mounting plane.

[0053] It should be noted that the least squares fitting criterion is a mathematical optimization technique. In this scheme, it is used to guide the SVD decomposition process, ensuring that the sum of squared errors between the extracted mounting plane of the rigid circuit board 32 and the 3D point cloud data is minimized, thereby obtaining the spatial pose that best fits the irregular curved surface in a physically meaningful way.

[0054] The right singular vector matrix (V matrix) is one of the three matrices output by Singular Value Decomposition (SVD). Its column vectors consist of eigenvectors of the covariance matrix of the input point cloud data, representing the principal component directions of the data distribution. In this scheme, it serves as the direct data source for solving the optimal normal vector.

[0055] The minimum singular value is the smallest value among the diagonal elements (singular values) of the diagonal matrix S after performing SVD decomposition on the decentralized point cloud, arranged in descending order. Geometrically, it corresponds to the direction in which the data points are most concentrated, i.e., the normal direction of the fitted plane.

[0056] The eigenvector (corresponding to the minimum singular value) is the column vector in the right singular vector matrix V that corresponds to the minimum singular value. Mathematically, this eigenvector defines a hyperplane (here, a plane), which physically represents the optimal normal vector of the mounting plane of the rigid circuit board 32.

[0057] A rotation transformation matrix is ​​a matrix used to describe the rotation of a three-dimensional spatial coordinate system. In this scheme, it is constructed based on the optimal normal vector and used to transform the three-dimensional point cloud data in the global coordinate system to a two-dimensional local coordinate system based on the plane of the rigid circuit board 32, thereby simplifying the spatial dimension.

[0058] In this embodiment, the point cloud data of the centerline 12 of the appearance surface 11 is extracted as a geometric reference by reading the three-dimensional model of the light-emitting device. The point cloud is mathematically analyzed by using the singular value decomposition (SVD) algorithm. The optimal fitting plane of the point cloud distribution is found according to the least squares criterion, and the normal of the plane is defined as the mounting reference of the rigid circuit board 32. Based on the normal vector, a rotation transformation matrix is ​​constructed to straighten the point cloud data that was originally in a complex three-dimensional space into a two-dimensional local coordinate system.

[0059] This process not only established the optimal tilt angle of the circuit board in physical space, but more importantly, through mathematical dimensionality reduction, it transformed the subsequent cumbersome three-dimensional optical optimization problem into a simple two-dimensional planar planning problem, laying the geometric foundation for automated lighting layout.

[0060] This embodiment uses the SVD algorithm for automatic optimization, which completely eliminates the attitude error caused by manual estimation and ensures that the circuit board is physically mounted as close as possible to the irregular curved surface, thereby minimizing the nonlinear fluctuation of the light mixing distance.

[0061] By introducing the rotation transformation matrix, the complex trigonometric function calculations in three-dimensional space are successfully avoided, reducing the problem to two-dimensional processing. This not only greatly improves the computational efficiency of the algorithm, but also provides great convenience for the construction of the light spot boundary equation and the solution of non-equidistant coordinates in subsequent steps.

[0062] This spatial mapping method based on mathematical models ensures the consistency and repeatability of each design result, fundamentally solving the boundary interference and ghosting errors that are easily caused by manual arrangement in multiple quadrants, and significantly improving the R&D accuracy and efficiency of high-end irregularly shaped light-emitting products.

[0063] In some embodiments, obtaining the right singular vector matrix based on point cloud data includes: calculating the centroid of the point cloud data; and performing singular value decomposition on the centroid-decentered point cloud data to obtain the right singular vector matrix.

[0064] It should be noted that the centroid refers to the arithmetic mean of all spatial coordinate points 31 in the point cloud data, that is, the geometric center of the point cloud.

[0065] Centroid removal is a data preprocessing operation. It involves subtracting the centroid coordinates from each point cloud coordinate, resulting in a new point cloud dataset centered at the origin of the coordinate system. This step eliminates the influence of the point cloud's position (translation) in space, preserving only its shape features.

[0066] Singular Value Decomposition (SVD) is a matrix factorization algorithm widely used in linear algebra. For any given m×n matrix A (here, the centroid-free point cloud matrix), SVD can decompose it into the product of three special matrices.

[0067] In this embodiment, the point cloud data of the centerline 12 of the input appearance surface 11 is statistically analyzed to calculate the centroid coordinates that can represent the geometric center of the trajectory. A "centroid removal" operation is performed to translate all point cloud coordinates so that their geometric centers coincide with the origin of the global coordinate system, thereby removing the translation attribute of the data and retaining only its inherent geometric configuration information. Singular value decomposition (SVD) is performed on the centroid-removed point cloud matrix to parse out the right singular vector matrix (V).

[0068] The column vectors of the right singular vector matrix represent the principal component directions of the point cloud data distribution. The column vector corresponding to the minimum singular value represents the direction of the minimum data fluctuation, which is mathematically defined as the normal direction of the fitting plane, thus providing the optimal spatial attitude reference for the rigid circuit board 32.

[0069] This embodiment effectively eliminates the interference of the absolute position of the point cloud on the calculation of the orientation vector by introducing a "centroid removal" preprocessing, ensuring that the extracted normal vector is unique and accurate regardless of how the 3D model is positioned in the software, greatly improving the robustness of the algorithm. Simultaneously, by directly extracting the eigenvector corresponding to the minimum singular value using SVD decomposition as the normal vector, it mathematically guarantees that the sum of squared errors between the obtained rigid circuit board 32 mounting plane and the point cloud trajectory is minimized (i.e., optimal fitting). This provides the most accurate spatial reference for minimizing the light mixing distance fluctuation and simplifying optical calculations in subsequent steps, fundamentally solving the problem of attitude optimization for rigid board mounting on irregular curved surfaces.

[0070] In some embodiments, determining the vertical mixing distance and the non-equidistant coordinates of the LEDs 33 in a single quadrant based on a two-dimensional local coordinate system and the optimal normal vector includes: calculating the vertical mixing distance of each LED 33 in the two-dimensional local coordinate system according to the optimal normal vector; establishing a light spot boundary constraint equation based on the Lambertian light emission characteristics and vertical mixing distance of the LEDs 33; and solving the light spot boundary constraint equation with the light spots of adjacent LEDs 33 connecting end to end as the constraint target to obtain the optimal total emission angle and the non-equidistant coordinates of the LEDs 33 in a single quadrant.

[0071] It should be noted that the luminous properties of the Lambert body refer to the inherent optical properties of the lamp bead 33 used in this application. Its luminous intensity decreases with the cosine of the angle between the emission direction and its own normal vector. That is, the light intensity is highest in the normal direction and gradually decreases in the oblique direction. The light intensity distribution in the half space has a rotational symmetry characteristic and there is no additional secondary optical lens interference.

[0072] The luminous half-angle refers to the angle between the emission direction and the normal direction when the emitted light intensity of the LED 33 drops to 50% of the peak light intensity in the normal direction.

[0073] The light spot boundary constraint equation is an equation based on the light spot coverage radius of a single LED bead 33 and the arc length distance between adjacent LED beads 33 on the lighting trajectory. It is used to describe the connection conditions of the light spot edges of adjacent LED beads 33.

[0074] The optimal total emission angle refers to the globally unique total emission angle obtained by nonlinear iterative solution under the premise of satisfying the constraint that all adjacent light spots are connected end to end.

[0075] The constraint of the light spot being connected end to end refers to the constraint condition that the light spot coverage areas of two adjacent LED beads 33 do not overlap or have gaps, and the outer boundary of the light spot of the previous LED bead 33 completely coincides with the inner boundary of the light spot of the next LED bead 33.

[0076] In this embodiment, the vertical mixing distance from each LED bead 33 to the irregularly shaped surface 11 is calculated along the optimal normal vector direction in the constructed two-dimensional local coordinate system. Then, based on the Lambertian light emission characteristics of the LED bead 33, its emission half-angle is determined. Combined with the vertical mixing distance, the light spot coverage radius of a single LED bead 33 on the surface 11 is derived. Based on this, the constraint equation of the light spot boundary of adjacent LED beads 33 is established. Finally, with the light spot end to end connected as the global constraint target, the optimal total emission angle is obtained by traversing and solving through a nonlinear iterative algorithm. Then, the non-equal spacing coordinates of the LED beads 33 on the rigid circuit board 32 in a single quadrant are obtained by reverse calculation, realizing the seamless splicing of light spots from a geometric level.

[0077] This embodiment automatically adjusts the spacing of the LED beads 33 by linking the vertical mixing distance with the luminous characteristics of the Lambertian light source, thus eliminating bright spots caused by overlapping light spots and dark areas caused by gaps from the source of geometric arrangement. This eliminates the high material and assembly costs of traditional solutions that rely on secondary lenses and light guide plates for light uniformity. At the same time, the fully automatic solution of non-equidistant coordinates replaces manual trial and error adjustment of each bead, completely avoiding interference errors at the intersection of multiple quadrants, improving the design efficiency of single-quadrant lighting, and providing a precise geometric reference for subsequent electrical closed-loop compensation, ensuring the high uniformity of the final light output.

[0078] In some embodiments, a light spot boundary constraint equation is established based on the Lambertian light emission characteristics of the LED 33 and the vertical mixing distance, including: determining the half-angle of the LED 33 based on the Lambertian light emission characteristics of the LED 33; and establishing the light spot boundary constraint equation based on the effective illumination boundary coordinates of the light spot, the coordinates of the direct projection point, the trajectory arc length between adjacent LEDs 33, the half-angle, and the vertical mixing distance.

[0079] It should be noted that the effective illumination boundary coordinates of the light spot refer to the edge position coordinates of the effective illuminance area formed by the light emitted by a single LED bead 33 on the irregularly shaped surface 11. In this scheme, this boundary is usually taken as the contour line where the light intensity drops to 50% of the peak value under the Lambertian light emission characteristics, representing a uniform light emission edge that is acceptable to the naked eye.

[0080] The coordinates of the direct projection point refer to the coordinates of the perpendicular intersection of the center normal of the LED bead 33 and the irregularly shaped surface 11. In the two-dimensional local coordinate system, this point is located on the centerline 12 of the surface 11, which is the endpoint for calculating the vertical light mixing distance and also the reference point for defining the center position of the light spot.

[0081] The arc length of the trajectory between adjacent LED beads 33 refers to the actual curve length along the centerline trajectory (i.e., the lighting path) of the outer surface 11, from the coordinates of the direct projection point of the previous LED bead 33 to the coordinates of the direct projection point of the next LED bead 33. This arc length is the core geometric link for establishing the boundary constraint equation of the light spot, and is different from the straight-line distance on the two-dimensional plane.

[0082] In this embodiment, the half-angle of the LED bead 33 is determined based on its physical characteristics, thereby defining the coverage radius of the light spot of a single LED bead 33. In the two-dimensional local coordinate system, the direct projection point of each LED bead 33 is extracted as the center of the light spot. Combined with the vertical mixing distance of this point, the effective illumination boundary coordinates of the light spot on the appearance surface 11 are calculated. The actual arc length of the trajectory along the curved surface between adjacent LED beads 33 is introduced as a constraint variable, and an equation is established to ensure that the outer boundary coordinates of the light spot of the previous LED bead 33 and the inner boundary coordinates of the light spot of the next LED bead 33 are precisely coincident on the trajectory, thereby locking the ideal state of the light spots being connected end to end at the geometric level.

[0083] This embodiment overcomes the physical limitations of traditional equidistant lighting layouts, which inevitably result in overlapping light spots (leading to bright stripes) or gaps (leading to dark stripes) on undulating surfaces, by establishing boundary constraint equations based on trajectory arc length and vertical mixing distance.

[0084] This precise constraint based on the physical optical model enables a continuous and uninterrupted wall-washing effect on complex three-dimensional curved surfaces using only LED beads 33, eliminating the generation of light and dark blemishes from the source. It also provides a unique and definite mathematical basis for solving non-equidistant coordinates in subsequent steps, greatly improving the fault tolerance rate of optical design and product yield.

[0085] In some embodiments, taking the alignment of the light spots of adjacent LED beads 33 as a constraint target, the boundary constraint equation of the light spots is solved to obtain the optimal total emission angle and the non-equidistant coordinates of the LED beads 33 in a single quadrant. This includes: obtaining the optimal total emission angle through a nonlinear iterative algorithm; and calculating the non-equidistant coordinates based on the optimal total emission angle.

[0086] It should be noted that the nonlinear iterative algorithm refers to a numerical calculation method used to solve overdetermined nonlinear equations with trigonometric function terms. It uses least squares as the optimization objective to find the globally optimal total emission angle that minimizes the sum of squared errors in the overlap of the light spot boundaries of all adjacent LED beads 33. Given the lack of analytical solutions to the constraint equations, the algorithm uses the inherent half-angle of the LED bead 33 as the initial iteration value, substitutes it into the light spot boundary constraint equation to calculate the total error. If the total error exceeds a pre-set convergence threshold (e.g., 0.01 mm, the upper limit of the positional error allowed by the process), the calculated value of the total emission angle is adjusted by a preset step size and the verification is repeated until the total error converges within the threshold, outputting the optimal solution. In this scheme, numerical algorithms such as the fzero function and Newton's iteration method can be used to implement this process.

[0087] In this embodiment, after establishing the boundary constraint equation for the light spot, since the constraints of all adjacent LED beads 33 share the same total emission angle variable, and the equation contains the tangent function term of the total emission angle, which belongs to an overdetermined nonlinear equation system without analytical solutions, the above-mentioned nonlinear iterative algorithm is introduced to solve the globally optimal total emission angle with least squares as the optimization objective. After obtaining the optimal total emission angle, it is substituted into the calculation formula of the light spot coverage radius of a single LED bead 33 to obtain the light spot coverage radius corresponding to each LED bead 33. Then, combined with the known trajectory arc length constraint between adjacent LED beads 33, the X / Y coordinates of each LED bead 33 in the two-dimensional local coordinate system are deduced, and finally, the non-equally spaced arrangement coordinates that perfectly adapt to the surface undulations are obtained.

[0088] This embodiment uses a nonlinear iterative algorithm with the least squares criterion to solve for the globally optimal total emission angle that minimizes the sum of the overlap errors of all light spots along the entire lighting trajectory. This avoids the defects such as overlapping bright spots and gaps and dark areas that are prone to occur in traditional equidistant lighting from a geometric perspective. It can achieve a continuous and uniform wall washing effect without relying on light-uniforming structures such as secondary lenses and light guide plates, and significantly reduces BOM costs.

[0089] The non-equidistant coordinates obtained by back-calculation of the optimal total luminous angle are perfectly adapted to the undulation characteristics of the curved surface, replacing the inefficient trial-and-error mode of manually adjusting the position of each lamp one by one. The time taken for single-quadrant lamp layout design is reduced from several hours to several minutes, improving efficiency. At the same time, it eliminates random errors of manual operation, ensures a high degree of consistency of optical effect during mass production, and significantly reduces the defect rate of mass production.

[0090] In some embodiments, the duty cycle of the pulse width modulation signal driving the LED 33 in a single quadrant is obtained based on the optimal normal vector and the vertical mixing distance, including: calculating the incident angle of the emitted light from the LED 33 relative to the optimal normal vector based on the Lambertian illuminance model, the optimal normal vector, non-equidistant coordinates, and the vertical mixing distance; calculating the Lambertian correction coefficient based on the incident angle; and calculating the duty cycle of the pulse width modulation signal driving the LED 33 based on the vertical mixing distance and the Lambertian correction coefficient.

[0091] It should be noted that the Lambertian illuminance model is a physical formula describing the illuminance distribution on a plane illuminated by a Lambertian light source. This Lambertian illuminance model quantifies the attenuation of light intensity as the incident angle increases.

[0092] The incident angle refers to the angle between the emitted light from the LED bead 33 and the optimal normal vector (i.e., the normal of the rigid circuit board 32 / the Z-axis of the local coordinate system). In the two-dimensional local coordinate system, it is calculated through geometric relationships using the non-equidistant coordinates of the LED bead 33 and the vertical mixing distance, and it characterizes the degree of tilt of the light illuminating the outer surface 11.

[0093] The Lambert correction factor is a quantitative compensation factor based on the incident angle for the light intensity attenuation of the Lambert body. It is used to incorporate the light intensity loss caused by oblique illumination into the compensation calculation, ensuring that the final output brightness of the oblique illumination area is consistent with that of the orthogonal illumination area.

[0094] The inverse square law is a fundamental physical law in the field of lighting, which states that under point light source illumination, the illuminance of an illuminated surface is inversely proportional to the square of the distance from the light source to the illuminated surface.

[0095] This embodiment relies on the constructed two-dimensional local coordinate system, combines the non-equidistant coordinates of each LED bead 33 with the vertical mixing distance, calculates the incident angle of its outgoing light relative to the optimal normal vector through geometric trigonometric functions, and then substitutes it into the Lambertian volume illuminance model to calculate the Lambertian correction coefficient of the corresponding LED bead 33, quantifies the light intensity attenuation ratio caused by oblique incidence, integrates the inverse square law, calculates the PWM duty cycle, and the vertical mixing distance and the Lambertian correction coefficient work together to transform the optical uniformity requirement into discrete drive signals of the underlying microcontroller.

[0096] This embodiment completely eliminates the oblique dark area problem that cannot be solved by simple distance compensation through dual compensation of the Lambertian correction factor and the inverse square law, keeping the illuminance deviation of the irregular surface 11 within a very small range and achieving a uniform wall washing effect close to that of an ideal Lambertian body. The complex optical homogenization requirements are completely converted into electrical drive parameters, eliminating expensive optical components such as secondary lenses, light guide plates, and light-blocking structures, significantly reducing BOM costs and assembly time. The PWM duty cycle lookup table matrix can be directly programmed into the microcontroller for execution, eliminating the need for manual brightness adjustment of each chip. This ensures optical consistency in mass production and reduces the optical debugging cycle of a single product from several days to minutes, significantly improving R&D and mass production efficiency.

[0097] In some embodiments, the duty cycle of the pulse width modulation signal driving the LED 33 is calculated based on the vertical mixing distance and the Lambertian correction factor, including: traversing the vertical mixing distance in a single quadrant to obtain the maximum vertical mixing distance; and calculating the duty cycle based on the vertical mixing distance, the maximum vertical mixing distance, and the Lambertian correction factor.

[0098] It should be noted that the maximum vertical mixing distance refers to the largest distance value extracted after traversing the vertical mixing distances of all LEDs 33 along the optimal normal vector to the outer surface 11 when calculating the single-quadrant LED 33. This parameter serves as a normalization benchmark to eliminate the influence of dimensions and ensure that the calculated PWM duty cycle falls within the 0%-100% range that the microcontroller can recognize.

[0099] Normalized distance squared is calculated by comparing the square of the vertical mixing distance of each LED (33) with the square of the maximum vertical mixing distance. This coefficient is an engineering implementation of the inverse square law, used to quantify the relative light attenuation caused by distance differences.

[0100] This embodiment iterates through the vertical mixing distances of all LEDs 33 within a single quadrant, selecting the maximum value as a reference. Then, for each LED 33, the ratio of its squared distance to the squared maximum distance (i.e., normalized squared distance) is calculated. This ratio reflects the degree of light attenuation caused by distance between the LED 33 and the farthest LED 33. Finally, this ratio is multiplied by the Lambertian correction factor and converted to a percentage to obtain the final PWM duty cycle. Through this mathematical mapping, the geometric distance and optical attenuation in the physical world are accurately converted into an electrical drive signal.

[0101] This embodiment uses the maximum vertical mixing distance as a normalization benchmark to ensure that the farthest LED 33 is driven with higher power, while the near-end LED 33 automatically reduces power according to the inverse square law, perfectly compensating for the illuminance difference caused by distance from an electrical perspective. Simultaneously, a Lambert correction factor is introduced to eliminate cosine attenuation caused by oblique illumination; the combined effect of these two factors ensures highly consistent illuminance across the surface 11. Furthermore, the generated duty cycle lookup table matrix can be directly programmed into a low-end microcontroller for execution, eliminating the need for expensive constant current control chips or complex analog dimming circuits. While ensuring top-tier optical quality, this significantly reduces material costs and firmware development difficulty, achieving highly efficient synergy between optics, mechanics, and electronics.

[0102] In some embodiments, the non-equidistant coordinates and duty cycles in a single quadrant are globally mirrored to obtain a spatial coordinate file and a duty cycle lookup table matrix representing the LEDs 33 in all quadrants. This includes: globally mirroring the non-equidistant coordinates and duty cycles in a single quadrant to obtain a spatial coordinate file and a duty cycle lookup table matrix representing the LEDs 33 in all quadrants. The non-equidistant coordinates and duty cycles of the LEDs 33 in all quadrants are output as a spatial coordinate file and a duty cycle lookup table matrix. The spatial coordinate file is used to determine the structural layout of the LEDs 33 on the rigid circuit board 32, and the duty cycle lookup table matrix is ​​used to determine the duty cycle of the pulse width modulation signal driving the LEDs 33.

[0103] It should be noted that global mirroring refers to the mathematical operation of using the inherent geometric symmetry axis of the irregular speaker's appearance (usually the X=0 and Y=0 planes) to copy and map the non-equally spaced coordinates and duty cycle parameters calculated in a single quadrant to the other three quadrants according to the spatial reflection transformation rules.

[0104] A spatial coordinate file is a structured file (such as a CSV, DXF, or Gerber coordinate file) that records the X, Y, and Z three-dimensional coordinate data of all LEDs 33 in the local coordinate system of the rigid circuit board 32. This file can be directly imported into CAD software for PCB-level routing or into a pick-and-place machine program for automated SMT placement.

[0105] The duty cycle lookup table matrix is ​​a two-dimensional array or list that stores the PWM duty cycle values ​​corresponding to all LEDs in all quadrants. This matrix is ​​stored in the microcontroller's flash memory as part of the firmware code. During runtime, the values ​​in this matrix are retrieved sequentially to configure the timer register, thereby outputting a precise pulse width modulation signal.

[0106] The algorithm identifies the geometric symmetry center of the product model, using the non-equidistant coordinates and PWM duty cycle obtained through SVD fitting and spot optimization within a single quadrant as source data. Based on the symmetry planes of X=0 and Y=0, the source data undergoes spatial coordinate inversion (±X,±Y) and direct duty cycle copying operations to automatically generate spatial arrangement data of LED beads 33 covering the entire quadrant. The system exports the global coordinates into an industry-standard spatial coordinate file and packages the global duty cycle into a lookup table matrix. The former guides the physical processing and mounting of the rigid circuit board 32, while the latter is directly burned into the microcontroller firmware, completing a seamless mapping from mathematical algorithms to physical entities.

[0107] On the design side, global mirroring technology allows engineers to focus on precise optical calculations in a single quadrant to automatically obtain a lighting scheme for the entire circumference, reducing the manual symmetry adjustment work that originally required several days to millisecond-level algorithm calculations and improving design efficiency.

[0108] In terms of production and optical performance, since the data for all quadrants originates from the same set of rigorously optimized source data, the cumulative errors that may be introduced by artificial mirroring and the defects of bright and dark lines at the boundaries of quadrants are completely eliminated, ensuring the absolute symmetry and visual harmony of the product in spatial optical output, and greatly improving the yield and mass production consistency of high-end irregular light-emitting products.

[0109] This embodiment also provides a light-emitting device, which is obtained by the above-described lamp bead layout method.

[0110] It is understandable that, since the light-emitting device provided in this embodiment is obtained through the above-described lamp bead placement method, it can also establish the spatial reference of the rigid circuit board 32 by utilizing point cloud data and the optimal normal vector, effectively resolving the assembly interference and height difference contradiction between the irregular appearance surface 11 and the rigid circuit board 32, which leads to obvious light and dark mottles on the light-emitting surface. On this basis, through the linkage calculation of non-equidistant coordinate arrangement and pulse width modulation duty cycle, the algorithm alone achieves accurate compensation for the light decay and distance difference of the Lambertian volume, eliminating the need for expensive secondary lenses and light guide structures in traditional solutions. At the same time, by using global mirroring technology, the precise calculation results of a single quadrant are extended to the entire domain without loss, which not only multiplies the design efficiency but also ensures the high symmetry and consistency of the optical output, ultimately achieving a high-quality uniform light emission effect with low material cost.

[0111] It should be noted that the light-emitting devices include, but are not limited to, speakers, smart home ambient light strips, automotive interior ambient lighting systems, building facade wall washer lights, ambient lights for curved home appliances (such as refrigerators and air conditioners), and ring fill lights for VR / AR headsets. Any device with a three-dimensional irregular appearance surface 11 and a built-in rigid circuit board 32 can achieve lens-free, low-cost, uniform light emission from curved surfaces on an inflexible rigid board.

[0112] Taking a speaker as an example, combined with the attached... Figures 2 to 14 The specific implementation process of this plan is described in detail below: like Figure 2 and Figure 3 As shown, the luminous speaker of this embodiment mainly includes a light structure 100, which includes an outer lamp housing 10, a bracket 20, and a light-emitting component 30. The outermost layer of the outer lamp housing 10 is a three-dimensional twisted free-form surface, namely the outer surface 11. This outer surface 11 faces away from the inner light-emitting component 30 and serves as the light-emitting surface for the wall-washing effect.

[0113] like Figure 3 As shown in the exploded view, the light-emitting component 30 consists of a rigid circuit board 32 and a plurality of LED beads 33 soldered thereon. The rigid circuit board 32 is fixed to the inner bottom of the outer lamp housing 10 by a bracket 20. In this structure, the rigid circuit board 32 is exemplarily made of a non-bendable planar material (e.g., FR4), while the LED beads 33 are bare light-emitting devices (e.g., LEDs) without secondary optical lenses, light-blocking structures, or light-guiding structures.

[0114] To achieve precise lighting placement on irregular curved surfaces, a geometric baseline needs to be defined. For example... Figure 4 and Figure 5 As shown, the center trajectory of the strip-shaped luminous area on the surface 11 is defined as the centerline 12 of the surface 11. Figure 7As shown, the centerline 12 is represented as a spatial curve in the 3D model. For digital processing, the centerline 12 is discretely sampled to obtain a series of lighting reference points 13, such as... Figure 8 As shown in the figure, these reference points form the basis of the point cloud data processed by subsequent algorithms.

[0115] In terms of optical characteristics, due to the use of bare LEDs, their luminous characteristics strictly follow the Lambertian distribution. For example... Figure 6 As shown, the light emitted by LED bead 33 is axially symmetrical about the normal of LED bead 33. Combined with... Figure 12 , Figure 13 and Figure 14 As shown, in the cross-section, the beam is symmetrical about the normal, with the left emission boundary defined as the first emission angle 34° and the right emission boundary as the second emission angle 35°. Physical properties dictate that the first emission angle 34° and the second emission angle 35° are strictly equal in their natural state. This geometrical optical constraint is a prerequisite for subsequently establishing the beam spot boundary equation and achieving lensless uniform light.

[0116] Based on the above structure, the system performs spatial attitude fitting and coordinate system dimensionality reduction. The system reads the three-dimensional coordinates of the centerline 12 of the appearance surface 11 and its lighting reference point 13, calculates its centroid, and performs singular value decomposition on the centroid-free point cloud. The eigenvector corresponding to the minimum singular value is extracted as the optimal spatial normal vector of the plane where the rigid circuit board 32 is located. Based on this normal vector, a rotation transformation matrix is ​​constructed to transform the three-dimensional point cloud data in the global coordinate system to the two-dimensional local coordinate system. This process logically realizes the mapping from "curved trajectory" to "planar arrangement", determines the optimal mounting tilt angle of the circuit board, and thus minimizes the nonlinear fluctuation of the light mixing distance.

[0117] The formula for singular value decomposition is 1-1, as follows: (1-1) Among them, P global It is the original 3D point cloud data matrix of the centerline 12 of the appearance surface 11 in the global coordinate system, with a dimension of N×3 (N is the number of sampling points). Each row represents the spatial coordinates of a sampling point on the X, Y, Z axes.

[0118] P center It is the centroid (geometric center) coordinate vector of the point cloud data.

[0119] U is a left singular vector matrix (the row space orthogonal basis of matrix A), with dimensions N×3, representing the projection distribution of the centroid-decentered point cloud along the principal component directions.

[0120] S is a singular value diagonal matrix with a dimension of 3×3. The diagonal elements s1, s2, and s3 are non-negative real numbers arranged in descending order (s1≥s2≥s3). The numerical values ​​represent the degree of dispersion or fluctuation of the point cloud data in the direction of the corresponding principal component.

[0121] V is a right singular vector matrix (an orthogonal basis of the column space of matrix A), with dimensions 3×3. Its column vectors v1, v2, and v3 correspond to the principal component directions of the singular values ​​s1, s2, and s3, respectively. The eigenvector v3 corresponding to the minimum singular value (s3) is the optimal normal vector of the mounting plane of the rigid circuit board 32. .

[0122] The point cloud data is exemplarily shown in Table 1, as follows:

[0123] In a two-dimensional local coordinate system, the system performs spot optimization and non-uniformly spaced coordinate calculation. For example... Figure 9 As shown, the system calculates the vertical mixing distance of each LED in the local coordinate system and establishes the light spot boundary constraint equation based on the Lambertian volume properties. With the goal of adjacent light spots meeting end-to-end, a nonlinear iterative algorithm is used to obtain the globally optimal total emission angle, which is 64.4° in this case. Based on this angle, the non-equidistant coordinates of the LED beads 33 in a single quadrant on the rigid circuit board 32 are calculated, as shown below. Figure 10 As shown, this non-equidistant arrangement effectively avoids the overlapping of light spots or dark areas that occur when projecting onto curved surfaces using traditional equidistant schemes.

[0124] Equation 1-2 of the light spot boundary constraint equation is shown below: (1-2) Among them, (X) Ai ,Y Ai (X) represents the effective illumination boundary coordinates of the light spot of the i-th LED bead 33 on the outer surface 11. Bi ,Y Bi ) represents the coordinates of the direct projection point, OD i This refers to the vertical mixing height of the LED bead 33.

[0125] Subsequently, the system performs closed-loop electrical compensation matrix derivation. Due to the undulations on the surface 11 causing varying vertical mixing distances at different points, the system calculates the incident angle of the emitted light from each LED 33 and the corresponding Lambertian correction coefficient based on the Lambertian volume illuminance model. Combining the inverse square law, the PWM duty cycle of each LED 33 is derived, as follows: Figure 11 As shown. This duty cycle lookup table matrix will be directly sent to the microcontroller to drive the LEDs 33 on the rigid circuit board 32, thereby achieving uniform illumination without the need for a lens.

[0126] The calculation formulas for the illuminance model, 1-3, are shown below: (1-3) Where E represents the actual illuminance of the target surface. The value represents the luminous flux received per unit area after distance attenuation and angle correction, typically measured in lux.

[0127] I0 represents the initial intensity of LED 33. It refers to the original luminous intensity of the light source in the direction perpendicular to the light emission direction (i.e., the normal direction, when α=0), and serves as a reference value for calculation.

[0128] This represents the cosine attenuation term of the Lambertian form. It is used to correct the effect of the incident angle of light on the lighting effect. Here, α is the angle between the light's exit direction and the normal to the target surface, also known as the incident angle. m is an exponential factor related to the half-emission angle (when m=1, it is a standard Lambertian form).

[0129] The calculation formula for the inverse square law is shown in equation 1-4 below: (1-4) Where P%i represents the compensation percentage of the PWM duty cycle of the i-th LED bead 33. That is, the electrical control signal that the microcontroller needs to send, with a value range between 0 and 100%, is used to directly adjust the brightness of the LED bead 33.

[0130] This represents the square of the vertical mixing distance of the i-th LED (33). max(OD) 2 ) refers to the square of the global maximum mixing distance. That is, the square of the maximum vertical mixing distance among all the LEDs 33 to be calculated. This parameter is used as the normalization denominator to eliminate extreme distance differences caused by the undulations of the appearance surface 11.

[0131] This represents the Lambertian correction factor for the i-th LED (LED 33). It is calculated based on the emission angle α. This is used to compensate for the loss of effective luminous flux caused by oblique light.

[0132] The non-equidistant coordinates and PWM duty cycle (energy percentage) of the LEDs 33 within a single quadrant (first quadrant) are exemplarily shown in Table 2, as follows:

[0133] Finally, the product's geometric symmetry is used to output a global mirror image. Figure 10The diagram shows the non-equidistant coordinates and duty cycles of LED beads 33 within a single quadrant. After mirroring through a plane of symmetry with X=0 and / or Y=0, the result is as follows: Figure 11 The coordinates and duty cycles of the non-equidistant spacing of LED beads 33 in all quadrants are shown.

[0134] For example, the system calculates the non-equidistant coordinates and duty cycles of the 6 LEDs within a single quadrant, and then performs a global mirroring by inverting the coordinates. The final output includes a complete spatial coordinate file containing all 24 LEDs (33), for use in CAD structural layout and SMT placement, as well as a global PWM duty cycle lookup table matrix for microcontroller firmware programming, completing the entire automated lighting layout and compensation process.

[0135] In some embodiments, the LED chip 33 is mounted on the rigid circuit board 32 according to the spatial coordinate file generated by the LED chip deployment method. The LED chip 33 determines the pulse width modulation signal driving the LED chip 33 based on the duty cycle lookup table matrix generated by the LED chip deployment method.

[0136] It should be noted that during the product manufacturing phase, the system imports the spatial coordinate file containing the three-dimensional positions of all quadrant LED beads 33 generated by the algorithm into CAD and SMT placement equipment. This guides the bare LED beads 33 to be precisely mounted onto the rigid, non-bendable circuit board 32 with non-uniformly spaced precision coordinates. Simultaneously, during the firmware burning phase, the duty cycle lookup table matrix is ​​written into the underlying microcontroller. When the system runs, the MCU sequentially retrieves values ​​from the matrix to configure timers, outputting PWM pulse signals that match the vertical mixing distance and Lambertian correction factor of each LED bead 33. Through differential dynamic compensation of electrical power, the light intensity attenuation caused by surface undulations and oblique illumination is offset, ultimately achieving highly uniform illuminance on the surface 11.

[0137] In summary, this application overcomes the physical limitations of the traditional "equal spacing without compensation" scheme by integrating opto-mechatronics. By introducing a PWM closed-loop compensation matrix that integrates vertical mixing distance and Lambertian correction coefficient, and through cross-verification by simulation and actual measurement, the illuminance deviation of the appearance surface 11 is controlled at an extremely low level, completely eliminating the phenomenon of light and dark dappling.

[0138] By using the SVD algorithm based on the least squares criterion, the mounting tilt angle with the minimum spatial fitting error is found for the inflexible rigid circuit board 32, which greatly reduces the difficulty of hardware engineering implementation of the three-dimensional distorted appearance without the need for flexible circuit boards or complex brackets 20.

[0139] By abandoning expensive secondary optical lenses and light guide structures, and by exploring the mathematical boundaries of the native Lambertian characteristics of bare LEDs, combined with non-equidistant coordinate optimization and electrical compensation, seamless splicing of light spots and wall washing effect are ensured while significantly reducing BOM costs, mold opening risks and production time.

[0140] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0141] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the technical solutions and core ideas of this application. Those skilled in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. These modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for controlling the placement of LED beads, characterized in that, The lamp bead placement method is applied to a light-emitting device, which includes a rigid circuit board and an external lamp housing with an external surface facing away from the rigid circuit board. The lamp bead placement method includes: Based on the three-dimensional model of the light-emitting device, the point cloud data of the centerline trajectory of the appearance surface, the optimal normal vector of the plane where the rigid circuit board is located, and the two-dimensional local coordinate system are determined. Based on the two-dimensional local coordinate system and the optimal normal vector, the vertical mixing distance and the non-equidistant coordinates of the LED beads in a single quadrant are determined. The duty cycle of the pulse width modulation signal driving the LED in the single quadrant is obtained based on the optimal normal vector and the vertical mixing distance. Globally mirror the non-equally spaced coordinates and duty cycles in the single quadrant to obtain a spatial coordinate file and a duty cycle lookup table matrix representing the LEDs in all quadrants.

2. The LED bead placement method according to claim 1, characterized in that, The determination of point cloud data of the centerline trajectory of the appearance surface, the optimal normal vector of the plane where the rigid circuit board is located, and the two-dimensional local coordinate system based on the three-dimensional model of the light-emitting device includes: The point cloud data of the centerline trajectory of the appearance surface is determined based on the three-dimensional model of the light-emitting device; The right singular vector matrix is ​​obtained based on the point cloud data; Based on the least squares fitting criterion, the eigenvector corresponding to the smallest singular value in the right singular vector matrix is ​​determined as the optimal normal vector of the mounting plane of the rigid circuit board. A rotation transformation matrix is ​​constructed based on the optimal normal vector to transform the point cloud data into a two-dimensional local coordinate system based on the mounting plane.

3. The LED bead placement method according to claim 2, characterized in that, The process of obtaining the right singular vector matrix based on the point cloud data includes: Calculate the centroid of the point cloud data based on the point cloud data; Singular value decomposition is performed on the point cloud data with the centroid removed to obtain the right singular vector matrix.

4. The LED bead placement method according to claim 1, characterized in that, The determination of the vertical mixing distance and the non-equidistant spacing coordinates of the LEDs in a single quadrant based on the two-dimensional local coordinate system and the optimal normal vector includes: In the two-dimensional local coordinate system, the vertical mixing distance of each LED bead is calculated based on the optimal normal vector; Based on the Lambertian light emission characteristics of the LED beads and the vertical light mixing distance, a light spot boundary constraint equation is established; Using the goal of ensuring that the light spots of adjacent LED beads are connected end to end, the boundary constraint equation of the light spots is solved to obtain the optimal total emission angle and the non-equidistant coordinates of the LED beads in a single quadrant.

5. The LED bead placement method according to claim 4, characterized in that, Based on the Lambertian light emission characteristics of the LED chips and the vertical mixing distance, a light spot boundary constraint equation is established, including: The half-angle of the LED is determined based on the Lambertian light-emitting characteristics of the LED. Based on the effective illumination boundary coordinates of the light spot, the coordinates of the direct projection point, the arc length of the trajectory between adjacent LEDs, the half-angle, and the vertical mixing distance, the boundary constraint equation of the light spot is established.

6. The LED bead placement method according to claim 5, characterized in that, The process of using the alignment of adjacent LED light spots end-to-end as a constraint objective, and solving the boundary constraint equations of the light spots to obtain the optimal total emission angle and the non-equidistant coordinates of the LEDs in a single quadrant, includes: The optimal total emission angle is obtained by solving a nonlinear iterative algorithm. The non-equidistant coordinates are obtained by back-calculation based on the optimal total emission angle.

7. The LED bead placement method according to claim 1, characterized in that, The step of obtaining the duty cycle of the pulse width modulation signal driving the LED in the single quadrant based on the optimal normal vector and the vertical mixing distance includes: Based on the Lambertian illuminance model, the optimal normal vector, the non-equidistant coordinates, and the vertical mixing distance, the incident angle of the emitted light from the LED relative to the optimal normal vector is calculated. Calculate the Lambert correction factor based on the incident angle; The duty cycle of the pulse width modulation signal driving the LED is calculated based on the vertical mixing distance and the Lambert correction factor.

8. The LED bead placement method according to claim 7, characterized in that, The step of calculating the duty cycle of the pulse width modulation signal driving the LED based on the vertical mixing distance and the Lambertian correction coefficient includes: Traverse the vertical mixing distances in the single quadrant to obtain the maximum vertical mixing distance; The duty cycle is calculated based on the vertical mixing distance, the maximum vertical mixing distance, and the Lambert correction factor.

9. The LED bead placement method according to any one of claims 1-8, characterized in that, The step of globally mirroring the non-equidistant coordinates and duty cycles in the single quadrant to obtain a spatial coordinate file representing the LEDs in all quadrants and a duty cycle lookup table matrix includes: Globally mirror the non-equally spaced coordinates and duty cycles in the single quadrant to obtain a spatial coordinate file and a duty cycle lookup table matrix representing the LEDs in all quadrants; The non-equal spacing coordinates and duty cycles of all LEDs in all quadrants are output as a spatial coordinate file and a duty cycle lookup table matrix. The spatial coordinate file is used to determine the structural layout of the LEDs on the rigid circuit board, and the duty cycle lookup table matrix is ​​used to determine the duty cycle of the pulse width modulation signal driving the LEDs.

10. A light-emitting device, characterized in that, The light-emitting device is obtained by the lamp bead placement method as described in any one of claims 1-9.

11. The light-emitting device according to claim 10, characterized in that, The LEDs are mounted on the rigid circuit board according to the spatial coordinate file generated by the LED deployment method; the LEDs determine the pulse width modulation signal to drive the LEDs according to the duty cycle lookup table matrix generated by the LED deployment method.

12. The light-emitting device according to claim 10, characterized in that, The rigid circuit board is mounted on the inside of the exterior lamp housing via a bracket.

13. The light-emitting device according to claim 12, characterized in that, The outer surface is the outermost three-dimensional twisted freeform surface of the lamp housing, and the lamp bead is a light-emitting device without at least one of the following: no secondary optical lens, no light-blocking structure, and no light-guiding structure.

Images