An indoor light therapy method based on dynamic balance of natural light and artificial lighting

Abstract

The present application relates to the technical field of indoor lighting, and relates to an indoor light therapy method based on dynamic balance of natural light and artificial lighting, which comprises the following steps: S1: predicting the values of each coefficient in the annual coefficient library-II at each time point in the day in a year based on a BP neural network; S2: calculating the values of each coefficient in the annual coefficient library-I at each time point in the day in a year based on a polynomial function group and in combination with each coefficient in the predicted annual coefficient library-II; S3: obtaining the pupil illuminance distribution of the indoor at each time point in the day in a year based on a polynomial function and in combination with each coefficient in the annual coefficient library-I; and S4: selecting an artificial light supplement scheme from a light supplement library based on a genetic algorithm and taking 1000-2000 lux as the indoor illuminance target range, so as to realize dynamic light supplement in a year. The present application can preferably realize indoor light therapy.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of indoor lighting, in particular to an indoor light therapy method based on dynamic balance of natural light and artificial lighting. BACKGROUND

[0002] Light therapy is a method of preventing and treating diseases by using sunlight, visible light and invisible light in artificial light source. Current light therapy mainly includes ultraviolet therapy, visible light therapy, infrared therapy and laser therapy.

[0003] Infrared rays acting on the human body mainly improve local blood circulation, promote swelling subsidence, analgesia, reduce muscle tension, relieve muscle spasm and dry exudative lesions.

[0004] When ultraviolet rays act on the human body, light energy causes a series of chemical reactions, has anti-inflammatory, analgesic and anti-rickets effects, and is commonly used to treat skin pyogenic inflammation and other dermatitis, pain syndromes, rickets or chondropathy, etc.; the ultraviolet rays with a wavelength of 310-313 nm are called narrow spectrum medium wave ultraviolet rays (NBUVB), which directly act on the skin affected area by concentrating the most biologically active part of the ultraviolet rays, while filtering out the harmful ultraviolet rays of the adverse wave band to the skin, having small side effects, acting on the cuticle of the skin, and having short onset time and fast effect. It has been widely used in hospitals for the treatment of diseases such as psoriasis, vitiligo, chronic eczema, neurodermatitis, atopic dermatitis, palmoplantar pustulosis, roseiform dermatitis, alopecia areata, parapsoriasis, chronic skin ulcer, and mycosis fungoides.

[0005] Visible light is the light that can be seen by the human eye. The method of treating diseases with visible light is visible light therapy. It mainly includes red light, blue light, blue-violet light and multi-spectrum therapy. Red light has excitatory effect; yellow light, green light and red light have opposite effects; blue-violet light can be used for the treatment of kernicterus.

[0006] Laser is a light generated by stimulated radiation light amplification, which has the characteristics of small divergence angle, good directionality, pure spectrum, good monochromaticity, high energy density, high brightness, and good coherence, and has thermal effect, mechanical effect and electromagnetic effect. It can be used for the diagnosis and treatment of many diseases.

[0007] However, the above light therapy cannot effectively combine natural light and artificial light according to indoor data. SUMMARY

[0008] The present application provides an indoor light therapy method based on dynamic balance of natural light and artificial lighting, which can overcome some or some defects of the prior art.

[0009] According to the indoor light therapy method based on dynamic balance of natural light and artificial lighting, the method comprises the following steps:

[0010] S1: Based on the BP neural network, the values of each coefficient in the annual coefficient library-Ⅱ at each time point during the day are predicted;

[0011] S2: Based on the polynomial function group, the values of each coefficient in the annual coefficient library-Ⅰ at each time point during the day are calculated by combining the predicted annual coefficient library-Ⅱ;

[0012] S3: Based on the polynomial function, the indoor pupil illuminance distribution at each time point during the day is obtained by combining each coefficient in the annual coefficient library-Ⅰ;

[0013] S4: Based on the genetic algorithm, the artificial light supplement scheme is selected from the light supplement library (6) with 1000-2000 lux as the indoor point illuminance target range, to realize dynamic light supplement throughout the year.

[0014] As a preferred, in step S1, the BP neural network adopts a three-layer BP neural network model, including an input layer, a hidden layer and an output layer, the neurons between adjacent two layers are connected to each other, and there is a corresponding connection weight; the input parameters of the BP neural network include time, solar elevation angle and azimuth angle, and the output is the values of each coefficient in the annual coefficient library-Ⅱ at each time point during the day, including A a , B a , C a ; A b , B b , C b ; A c , B c , C c ; A d , B d , C d ; A e , B e , C e .

[0015] As a preferred, in steps S2 and S3, polynomial fitting is performed between pupil illuminance and depth y, 18 width steps and 13 polynomial function coefficients of 10 hours per day on average are calculated, and the output is coefficient I matrix; polynomial fitting is performed on coefficient I and width x, and coefficient II matrix is output; by comparing the prediction accuracy of different polynomial equations on pupil illuminance distribution, it is finally determined that one quartic polynomial and five quadratic polynomials have the highest prediction accuracy on indoor pupil illuminance distribution, and the fitting degree R 2 value>0.999;

[0016] The polynomial formula is:

[0017] Ecor=A*y 4 +B*y 3 +C*y 2 +D*y+E;

[0018] A = A a *x 2 + B a *x + C a ;

[0019] B = A b *x 2 + B b *x + C b ;

[0020] C = A c *x 2 + B c *x + C c ;

[0021] D = A d *x 2 + B d *x + C d ;

[0022] E = A e *x 2 + B e *x + C e .

[0023] As preferred, in step S4, the light supplement library is composed of hundreds of groups of light distribution data under different lighting conditions in the room, and the lighting conditions include:

[0024] a. Lighting grid, the node distance of the grid light is 0.2-0.5m, and the installation height is 3m;

[0025] b. Luminous flux of the light, the light distribution data of 2000lm, 4000lm, 6000lm and 8000lm lamps are included in the light supplement library, and the light distribution modes of the lamps include: lambert light distribution, light distribution scheme with 70°, 50°, 30° light cutting angle based on lambert light distribution;

[0026] c. Number of lighted lamps, according to the number of light deficiency throughout the year, the required number range of various light output lamps is calculated as follows:

[0027] 2000lm: 5-40, 4000lm: 5-30, 6000lm: 5-20, 8000lm: 5-15;

[0028] d. Lighted lamp row, light distribution data information of 1-5 rows of lighting mode is included;

[0029] e. Lateral arrangement mode, including lamps concentrated in the east area of the room, lamps concentrated in the middle area of the room and lamps concentrated in the west area of the room.

[0030] Preferably, in step S4, the steps of the genetic algorithm are as follows:

[0031] 1) Encode the supplementary lighting distribution data to be screened in an automatically sorted and non-repeating manner;

[0032] 2) Select the number of supplementary lighting distribution data points, set each data set to consist of the number of lamps, lamp output, lamp light distribution, and number of lamp rows, and generate initial data;

[0033] 3) Combine the supplementary lighting distribution data with the daylight data to calculate whether the target range of 1000-2000 lux for indoor illuminance is met;

[0034] 4) Replace the supplementary lighting data until the supplementary lighting distribution data meets the settings of step 3);

[0035] 5) Output the results and select the optimal artificial lighting scheme from the generated supplementary lighting distribution data.

[0036] This invention provides a dynamically stable indoor phototherapy method combining natural lighting and artificial lighting. The method is based on a BP neural network, multinomial functions, and a coefficient library. It combines real-time light climate data to predict indoor natural light distribution data, and then uses a genetic algorithm to dynamically select the optimal artificial lighting scheme, achieving an effective light intervention threshold (1000-2000 lux) across the entire indoor space, including all areas, all fields of view, and all time periods. The proposed method can be applied to therapeutic lighting in elderly care spaces, fully considering regional light climate and solar eclipse patterns. Combined with intelligent control, it achieves precise and directional lighting to ensure that the effective threshold is reached across all areas, all fields of view, and all time periods in the elderly care space. Attached Figure Description

[0037] Figure 1 This is a flowchart illustrating an indoor phototherapy method based on the dynamic balance between natural light and artificial lighting, as described in the embodiments.

[0038] Figure 2(a) shows A in the coefficient library-II in the embodiment. a B a C a A schematic diagram;

[0039] Figure 2(b) shows A in the coefficient library-II in the embodiment. b B b C b A schematic diagram;

[0040] Figure 2(c) shows A in the coefficient library-II in the embodiment. c B c C c A schematic diagram;

[0041] Figure 2(d) shows A in the coefficient library-II in the embodiment. d Bd , C d , C e , C e , C e , C a , C a , C a , C b , C b , C b , C c , C c , C c , C d , C

[0042] Figure 2 (e) is a schematic diagram of A

[0043] Figure 3 is a genetic algorithm logic diagram in the embodiment. DETAILED DESCRIPTION

[0044] For a further understanding of the present application, reference will be made to the following description of embodiments taken in conjunction with the accompanying drawings. It is to be understood that the embodiments are merely illustrative of the present application and should not be construed as limiting the present application in any way.

[0045] EMBODIMENT

[0046] As shown in Figure 1 , an indoor light therapy method based on dynamic balance of natural light and artificial lighting includes the following steps:

[0047] S1: Based on BP neural network 1, predict the values of each coefficient in the annual coefficient library-Ⅱ at each time point during the day throughout the year;

[0048] S2: Based on the polynomial function group 2, combine the predicted values of each coefficient in the annual coefficient library-Ⅱ to calculate the values of each coefficient in the annual coefficient library-Ⅰ at each time point during the day throughout the year;

[0049] S3: Based on the polynomial function 3, combine the values of each coefficient in the annual coefficient library-Ⅰ to obtain the pupil illuminance distribution 4 of the indoor at each time point during the day throughout the year;

[0050] S4: Based on the genetic algorithm 5, select the artificial light supplement scheme in the light supplement library (6) with 1000-2000 lux as the indoor light target range to achieve dynamic light supplement throughout the year.

[0051] In step S1, the BP neural network 1 adopts a three-layer BP neural network model, which includes an input layer, a hidden layer and an output layer. The neurons between adjacent two layers are connected to each other, and there is a corresponding connection weight. The input parameters of the BP neural network include time, solar elevation angle and azimuth angle, and the output is the values of each coefficient in the annual coefficient library-Ⅱ at each time point during the day throughout the year, including A a , B a , C a ; A b , B b , C b ; A c , B c , C c ; A d , Bd , C d ; A e , B e , C e .

[0052] In steps S2 and S3, a polynomial fitting is performed between the pupil illuminance and the depth y, 18 width steps and 13 average 10-hour-per-day polynomial function coefficients are calculated, and the output is the coefficient I matrix ([18x130], width 18 steps, average 10 hours per day, average number of days calculated 13 days); a polynomial fitting is performed on the coefficient I and the width x, and the coefficient II matrix ([1x130]) is output; by comparing the prediction accuracy of the pupil illuminance distribution by different polynomial equations, it is finally determined that a quartic polynomial and five quadratic polynomials have the highest prediction accuracy for the indoor pupil (Ecor) illuminance distribution, and the fitting degree R 2 value is >0.999;

[0053] The polynomial formula is:

[0054] Ecor=A*y 4 +B*y 3 +C*y 2 +D*y+E;

[0055] A=A a *x 2 +B a *x+C a ;

[0056] B=A b *x 2 +B b *x+C b ;

[0057] C=A c *x 2 +B c *x+C c ;

[0058] D=A d *x 2 +B d *x+C d ;

[0059] E=A e *x 2 +B e *x+C e .

[0060] The coefficient library-II is shown in FIG. 2(a), FIG. 2(b), FIG. 2(c), FIG. 2(d), and FIG. 2(e); the ordinate of FIG. 2(a), FIG. 2(b), FIG. 2(c), and FIG. 2(d) is the same as that of FIG. 2(e).

[0061] In step S4, the light supplement library 6 is composed of hundreds of groups of light distribution data under different indoor lighting conditions, including:

[0062] a. Lighting grid, the grid lamp node spacing is 0.2-0.5m, and the installation height is 3m;

[0063] b. Luminous flux of the lamp, the light supplement library contains light distribution data of 2000lm, 4000lm, 6000lm and 8000lm lamps, and the light distribution mode of the lamp includes: Lumenier light distribution, light distribution scheme with 70°, 50°, 30° and other cut-off angles based on Lumenier light distribution;

[0064] c. Number of lighted lamps, according to the number of light deficiency throughout the year, the required number range of various light output lamps is calculated as 5-40 (2000lm), 5-30 (4000lm), 5-20 (6000lm), and 5-15 (8000lm);

[0065] d. Lighted lamp row, containing light distribution data information of 1-5 rows of lighting arrangement;

[0066] e. Lateral arrangement, including lamps concentrated in the eastern region of the room, lamps concentrated in the middle region of the room, and lamps concentrated in the western region of the room.

[0067] As a preferred, in step S4, the steps of the genetic algorithm 5 are:

[0068] 1) Encode the to-be-screened light supplement distribution data in an automatic sorting and non-repeating manner;

[0069] 2) The number of light supplement distribution data to be selected is set, each group of data is composed of the number of lamps, lamp output (lm), lamp light distribution, and lamp row number, and initial data is generated;

[0070] 3) Combine the light supplement distribution data with the daylight data to calculate whether it meets the indoor each point illuminance target range of 1000-2000lux;

[0071] 4) Replace the light supplement data until the light supplement distribution data meets the setting of step 3);

[0072] 5) Output the result, select the optimal artificial light supplement scheme from the generated light supplement distribution data.

[0073] The genetic algorithm logic is as shown in Figure 3 . Figure 3The artificial light supplement scheme optimization strategy based on a genetic algorithm is shown. The light library contains a large number of light schemes and can be continuously expanded. Each light scheme has specific identity information, that is, the genetic gene of the genetic algorithm, including but not limited to the number of lamps, the output of the lamps (lm), the light distribution of the lamps, the number of rows of lamps, and the position of the outer edge of the lamps. These genes randomly generate several populations (that is, corresponding light scheme populations). For the natural lighting conditions at a certain time, the genetic algorithm generates a random light scheme population based on the above, respectively weighted with natural lighting, obtains the indoor pupil illuminance distribution under the simultaneous action of natural lighting and artificial light supplement, and compares the target pupil illuminance of 1000-2000 lux to obtain the possible population position of the best light supplement scheme. Through genetic, mutation, recombination, and crossover operators, a new light scheme population is obtained. The natural light is weighted and compared to select the best one. If the condition (1000-2000 lux) is met, it is a useful scheme. If the condition is not met, evolution (genetic, mutation, recombination, and crossover) is performed to obtain the next generation of light scheme population until the best light scheme is found.

[0074] The embodiment also provides an indoor light therapy system based on dynamic balance of natural light and artificial lighting, which adopts the indoor light therapy method based on dynamic balance of natural light and artificial lighting.

[0075]

[0076] a) Software part

[0077] Natural light prediction model:

[0078] The BP neural network model with a three-layer structure includes an input layer, four hidden layers, and an output layer. The neurons in adjacent layers are connected to each other, and there is a corresponding connection weight. The input parameters of the BP neural network include time, solar elevation angle, and azimuth angle. The output is the annual data of coefficient-Ⅱ, including A a , B a , C a ; A b , B b , C b ; A c , B c , C c ; A d , B d , C d ; A e , B e , C eThe values of each coefficient in the annual coefficient library-I at each time point in the daytime throughout the year are calculated in combination with each coefficient in the predicted annual coefficient library-II. In combination with each coefficient in the annual coefficient library-I, the prediction model of the pupil illuminance distribution of indoor natural light in a dynamic state throughout the year and throughout the day is obtained.

[0079] Natural light correction model:

[0080] In combination with the collected real-time light climate data, the BP neural network calculation is corrected.

[0081] Dynamic light supplement model:

[0082] Based on the genetic algorithm, the light library is formed according to the number of lamps, the output (lm) of the lamps, the light distribution of the lamps, and the lamp arrangement mode of the lamp rows. In the light library, the optimal artificial light supplement scheme is selected to achieve dynamic light supplement throughout the year. At the same time, the genetic algorithm will also select a better artificial light supplement scheme through updating and learning.

[0083] b) Hardware part

[0084] Server:

[0085] A set of high-speed devices for accepting and processing real-time information, including but not limited to computers, routers, network cables, etc. The internal installation of "natural light prediction model", "dynamic light supplement model", "light climate correction model" and other software for predicting natural light and dynamic light supplement.

[0086] Real-time environmental parameter acquisition device:

[0087] The device integrates various components to achieve comprehensive collection of environmental information to meet the needs of the system. The analysis and conversion method of related parameters is realized based on the previous research results. The device includes but is not limited to the following components: 1) time information acquisition device, which acquires real-time time information; 2) light environment parameter acquisition device, which acquires real-time light environment parameter information.

[0088] Ultra-high-speed information transmission and conversion device:

[0089] The device includes a display device and an operating lever. The light intervention device system calibration method is based on a large number of previous polynomial functions, BP networks, and genetic algorithms to obtain dynamic light supplement quantity and distribution calculation models throughout the year. Through the optimal selection of various light supplement libraries, the device is adjusted to improve the adaptability of the device.

[0090] Light intervention lamp assembly:

[0091] According to the algorithm, the optimal light supplement combination method is obtained in combination with natural light to obtain overall high illuminance and high uniformity. It is realized that the pupil illuminance of the elderly in the moving state is always maintained within the effective threshold range, and the indoor full area is covered.

[0092] The above description of the application and its embodiments is illustrative and not restrictive, and the embodiments shown in the drawings are only one of the embodiments of the application, and the actual structure is not limited thereto. Therefore, if a person skilled in the art is inspired by it, without departing from the purpose of the application, similar structural modes and embodiments can be designed without creativity, which should belong to the protection scope of the application.

Claims

1. A method of indoor light therapy based on dynamic balancing of natural light and artificial illumination, characterized in that: Comprise the following steps: S1: Based on BP neural network (1), predict the value of each coefficient in the annual coefficient library-Ⅱ in the annual day each time point; The BP neural network (1) adopts a three-layer structure BP neural network model, including an input layer, a hidden layer and an output layer, neurons between adjacent two layers are connected with each other, and have a corresponding connection weight; the input parameters of the BP neural network include time, solar elevation angle and azimuth angle, and the output is the numerical value of each coefficient in the annual coefficient library-Ⅱ at each time point in the day of each coefficient, including A a , B a , C a ; A b , B b , C b ; A c , B c , C c ; A d , B d , C d ; A e , B e , C e ; S2: Based on the polynomial function group (2), combined with the prediction of the annual coefficient library-Ⅱ in each coefficient, the value of each coefficient in the annual coefficient library-Ⅰ in the annual day each time point is calculated; S3: Based on the polynomial function (3), combined with the annual coefficient library-Ⅰ in each coefficient, the pupil illuminance distribution (4) in the indoor of each time point in the annual day is obtained; In steps S2 and S3, polynomial fitting is performed between the pupil illuminance and the depth y, 18 width steps and 13 average 10-hour-per-day polynomial function coefficients are calculated, and the output is a coefficient I matrix; polynomial fitting is performed on the coefficient I and the width x, and a coefficient II matrix is output; by comparing the prediction accuracy of different polynomial equations on the pupil illuminance distribution, it is finally determined that a quartic polynomial and five quadratic polynomials have the highest prediction accuracy on the indoor pupil illuminance distribution, and the fitting degree R 2 > 0.

999. Wherein the polynomial formula is: Ecor = A * y 4 + B * y 3 + C * y 2 + D * y + E; A = A a * x 2 + B a * x + C a ; B = A b x 2 + B b x + C b ; C = A c x 2 + B c x + C c ; D = A d * x 2 + B d * x + C d ; E = A e * x 2 + B e * x + C e ; Ecor is the indoor pupil illuminance; S4: Based on genetic algorithm (5), taking 1000-2000 lux as the indoor each point illuminance target range, the artificial light supplement scheme is selected in the light supplement library (6), and the dynamic light supplement in the whole year is realized.

2. The method of claim 1, wherein the method is based on dynamic balance between natural light and artificial lighting. In step S4, the light supplement library (6) is composed of hundreds of groups of light distribution data under different indoor lighting conditions, and the lighting conditions are: a, lighting grid, the node distance of grid lamp is 0.2-0.5 m, and the installation height is 3 m; b, light flux of lamp, the light distribution data of 2000 lm, 4000 lm, 6000 lm and 8000 lm four kinds of lamps are contained in the light supplement library, and the light distribution mode of lamp includes: lambert light distribution, light distribution scheme with 70°, 50°, 30° light cutting angle based on lambert light distribution; c, the number of bright lights, according to the number of light deficiency in the whole year, the required number range of various light output lamps is calculated as follows: 2000 lm: 5-40, 4000 lm: 5-30, 6000 lm: 5-20, 8000 lm: 5-15; d, the number of bright lights, containing the light distribution data information of 1-5 rows of lighting mode; e, horizontal arrangement mode, including lamp concentrated in the east area of the room, lamp concentrated in the middle area of the room and lamp concentrated in the west area of the room.

3. The method of claim 1, wherein the method further comprises: determining a time of day; and determining a time of year. In step S4, the steps of genetic algorithm (5) are: 1) encode the to-be-screened light supplement distribution data in an automatic sorting and non-repeating manner; 2) the number of selected light supplement distribution data is set, each group of data is composed of lamp number, lamp output, lamp light distribution and lamp array, and initial data is generated; 3) combine the light supplement distribution data with the daylight data to calculate whether it meets the 1000-2000 lux indoor each point illuminance target range; 4) replace the light supplement data until the light supplement distribution data meets the setting of step 3); 5) output results, select the optimal artificial light supplement scheme from the generated light supplement distribution data.

Images