Light source light mixing control method, intelligent lighting equipment and indoor lighting system

Abstract

A light source light mixing control method, an intelligent lighting device and an indoor lighting system, the control method mainly comprises the steps of light source selection, data acquisition and processing, BP neural network design and neural network training model, the method can be applied to the indoor lighting system, and the indoor lighting system can perform light mixing according to specific parameters such as the age of a user and the size of an indoor area. According to the intelligent lighting equipment adopting the control method, precise control over the light source and the color temperature is achieved, and therefore the lighting requirements of users of different ages in different scenes are met.

Description

Technical Field

[0001] The present invention relates to the technical field of light environment control, and particularly to a method for controlling light mixing of light sources, an intelligent lighting device, and an indoor lighting system. Background Art

[0002] With the progress of technology and the development of society, people's demand for lighting has changed from a simple requirement for brightness to the pursuit of health and comfort. As an emerging lighting concept, healthy lighting emphasizes the impact of lighting on human health, especially biological rhythms. Research shows that light not only affects visual perception but also affects people's physiological rhythms through non-visual pathways, such as the secretion of melatonin, sleep quality, and mood regulation. Therefore, it is particularly important to develop a lighting system that can regulate biological rhythms and promote health.

[0003] Existing lighting systems often ignore the differences in lighting requirements for different ages and environments and lack the functions of personalization and biological rhythm regulation. The sensitivity and requirements of children and the elderly to light are different from those of adults, but many existing lighting solutions do not take these differences into account. In addition, existing systems are rarely able to dynamically adjust lighting parameters according to environmental changes (such as time, season) and individual differences (such as age, health status). Summary of the Invention

[0004] Aiming at the deficiencies in the background art, the purpose of the present invention is to provide a method for controlling light mixing of light sources, an intelligent lighting device, and an indoor lighting system.

[0005] To achieve the above purpose, the present invention provides the following technical solutions:

[0006] A method for controlling light mixing of light sources, characterized by comprising the following steps:

[0007] S1. Light source selection: Select red, yellow, green, blue, cyan, and white as light sources;

[0008] S2. Data collection and processing: Collect data on the luminous flux, color temperature, color rendering index, and photobiological rhythm factor CAF of different age groups of the light sources, and process the collected data so that the data is normalized within the interval [0,1]; A S3. Design of BP neural network: including:

[0009] Input layer: There are 5 nodes, corresponding to luminous flux, color temperature, color rendering index, rhythm stimulation factor, and user age respectively;

[0010] Hidden layer: There are 20 base points;

[0011] Output layer: There is 1 node, corresponding to the optimal light mixing ratio;

[0012] The output layer has 6 nodes, corresponding to the duty cycles of the red, yellow, green, blue, cyan, and white light sources respectively;

[0013] The input layer and the output layer use activation functions to process the node data;

[0014] The mathematical expression of the activation function is:

[0015] S4. Train the neural network model: including the following steps:

[0016] S41. Collect the dataset samples described in S2;

[0017] S42. Divide the dataset collected in S41 into a training set, a validation set, and a test set;

[0018] S43. Apply weights and biases between the input layer, the hidden layer, and the output layer. The weight matrix from the input layer to the hidden layer is W1 ∈ R 20×5 , and the bias vector b1 ∈ R 20 ;

[0019] The weighted input of the hidden layer is z1 = W1 × x + b1, and the activation output of the hidden layer is a1 = σ(z1);

[0020] The weight matrix from the hidden layer to the output layer is W2 ∈ R 6×20 , and the bias vector b2 ∈ R 6 ;

[0021] The weighted input of the output layer is z2 = W2 × a1 + b2, and the activation output of the output layer is a2 = σ(z2);

[0022] S44. Introduce a formula for reducing errors: where e represents the magnitude of the error, n is the number of training times, y k ∈ R 6 is the predicted value of the output layer, is the true value of the output layer.

[0023] Furthermore, S2 also introduces the IQR method to identify and remove outliers from the collected data.

[0024] Furthermore, the collection of the photobiological rhythm factors collected in S2 includes the following steps:

[0025] S21. Calculate the spectral transmittance of the human eye at different ages. The following formula can be used when calculating the spectral transmittance:

[0026] τ(λ,A) = 10 -D(λ,A) ;

[0027] Among them, λ is the wavelength and A is the age.

[0028] S22. Introduce a correction factor M(A) for adjusting the pupil diameter during the collection process. The adjusted pupil diameter can be calculated using the following formula:

[0029] M(A) = [1 - c·(A - 25)] 2 ;

[0030] Among them, c is the adjustment coefficient for adjusting the through-hole diameter of different ages in the formula of S22, and c = 0.00559

[0031] S23. Use the non-visual spectral response curve C(λ) and the photopic spectral response curve V(λ) for the spectral power distribution P(λ) of the light source, the spectral transmittance τ(λ,A), the correction factor M(A), and the area Aret of the illuminated area on the retina, as shown in the appendix Figure 1 to calculate the circadian photoreception factor CAF based on different ages A , the circadian photoreception factor CAF based on different ages A The calculation formula is as follows:

[0032]

[0033] Furthermore, when using the activation function to process the data collected in S2, it needs to be initialized with random numbers. Among them, weight initialization: use the Xavier uniform distribution, range: [-0.447, 0.447]; bias initialization: set to 0 or a small random value close to 0.

[0034] Furthermore, the input data in S4 is the correlated color temperature CCT, luminous flux Φ, color rendering index Ra, circadian photoreception factor CAF A and age A;

[0035] The output result in S4 is the duty cycle of the red, yellow, green, blue, cyan, and white light sources, and each duty cycle is denoted as D R , D Y , D G , D B , D C , D WW .

[0036] An intelligent lighting device uses the light source mixing control method described above. The intelligent lighting device includes a circuit board, on which a light source module is provided. The light source module is electrically connected to a battery. The circuit board is also provided with a receiving module for receiving user input information, a control module for controlling the light source module, and a wireless signal transmission module. The light source module, the receiving module, the control module, the wireless signal transmission module, and the battery are all electrically connected.

[0037] Further, the light source module includes several LEDs that can emit red, yellow, green, blue, cyan, and white light sources. Among them, the LED that emits white light is located at the center of the light source module, and the LEDs that emit red, yellow, green, blue, and cyan light sources are distributed around the LED that emits white light, and all the LEDs are electrically connected to each other.

[0038] An indoor lighting system, characterized in that it includes the above-mentioned intelligent lighting device. The indoor lighting system includes a children's room lighting system and a study room lighting system. Users can input the user's age and room area parameters to the intelligent lighting device. The intelligent lighting device calculates and adjusts the luminous flux Φ based on the input parameters, so that the luminous flux Φ is maintained within the range of 150 lx to 500 lx.

[0039] Further, the children's room lighting system includes:

[0040] Daytime lighting mode: The intelligent lighting device adjusts the correlated color temperature CCT of the LED lamp beads to be maintained at the 5000K level, the color rendering index Ra is not less than 92, and the circadian action factor CAF of the environment A is maintained at the 0.9 level, and the luminous flux Φ is maintained at the 500 lx level; and

[0041] Nighttime lighting mode: The intelligent lighting device adjusts the correlated color temperature CCT of the LED lamp beads to be within the range of 2700K to 3000K, the color rendering index Ra is not less than 90, and the circadian action factor CAF of the environment A is less than 0.3, and the luminous flux Φ is maintained at the 150 lx level.

[0042] Further, the study room lighting system includes:

[0043] Teenager lighting mode: The correlated color temperature CCT is maintained at the 5000K level;

[0044] Adult lighting mode: The correlated color temperature CCT is within the range of 3000K to 5000K; and

[0045] Elderly lighting mode: The correlated color temperature CCT is maintained at the 4000K level.

[0046] The beneficial effects of the present invention are as follows:

[0047] 1. A method for controlling the mixing of light sources proposed by the present invention collects the luminous flux, color temperature, color rendering index of the light source, and the circadian photobiological factor CAF A data for different age groups, and inputs the collected data into a neural network model. By training the neural network model, the duty cycle of the selected six-color LED light source is accurately predicted and controlled, realizing specific light source parameter adjustment to meet the lighting needs of users of different age groups in different scenarios.

[0048] 2. A method for controlling the mixing of light sources proposed by the present invention introduces a neural network model into the control method. The BP neural network includes an input layer, a hidden layer, and an output layer. The input layer can receive specific parameters. The hidden layer extracts the abstract features of the data through weighted summation of weights and biases and the processing of activation functions. The output layer converts the features extracted by the hidden layer into the final prediction result. This clear structural design makes the calculation process of the model orderly and is also convenient for subsequent parameter optimization and model adjustment. At the same time, the neural network model divides the collected data points into a training set, a validation set, and a test set, which can make full use of data resources. The validation set plays a monitoring role during the model training process, and the test set is used to finally evaluate the performance of the model.

[0049] 3. A method for controlling the mixing of light sources proposed by the present invention accurately calculates the circadian photobiological factor CAF based on different ages by calculating the spectral transmittance and pupil diameter of the human eye for different age groups, and combining the spectral power distribution, spectral transmittance, correction factor, retinal illumination area of the light source, and the non-visual spectral response curve and the photopic spectral response curve. A Understanding the light requirements and responses of different age groups can provide a scientific basis for the design and improvement of lighting products.

[0050] 4. An intelligent lighting device proposed by the present invention includes a light source module, a receiving module, a control module, and a wireless signal transmission module. The receiving module receives the input information of the user, and intelligently adjusts the lighting parameters of the LED light source according to this information, and cooperates with the other modules to realize the integrated and intelligent control of the lighting device.

[0051] 5. An indoor lighting system proposed by the present invention. The indoor lighting system includes a children's room lighting system and a study lighting system. The children's room lighting system provides high-color-temperature and high-color-rendering-index lighting during the day and low-color-temperature and low-photobiological rhythm factor lighting at night to meet the lighting needs of children at different times. The study lighting system provides a customized lighting solution for users of different ages, taking into account the eye transmittance and pupil correction factor M value of the users, adjusting the color temperature and luminous flux to adapt to different learning and reading activities, and dynamically adjusting the lighting parameters to adapt to the biological rhythm of the occupants. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0053] Figure 1 It is a schematic diagram of a light source mixing control method;

[0054] Figure 2 It is a schematic diagram of the collection method for collecting photobiological rhythm factors in S2;

[0055] Figure 3 It is a schematic diagram of the method for training a neural network model in S4;

[0056] Figure 4 It is the non-visual spectral response curve C(λ) and the photopic curve V(λ);

[0057] Figure 5 It is the spectra of R\Y\G\B\C\W six-color LED light sources;

[0058] Figure 6 It is a neural network structure diagram;

[0059] Figure 7 It is a schematic diagram of an intelligent lighting device;

[0060] Figure 8 It is a schematic diagram of an indoor lighting system.

[0061] In the figure, 101 is the input layer; 102 is the hidden layer; 103 is the output layer; 20 is the intelligent lighting device; 201 is the circuit board; 202 is the light source module; 203 is the receiving module; 204 is the control module; 205 is the wireless signal transmission module; 206 is the battery. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0062] The following will combine Figures 1 to 8 to describe the present invention in detail.

[0063] A method for controlling the mixing of light sources, comprising the following steps:

[0064] S1. Light source selection: Select red, yellow, green, blue, cyan, and white as light sources. The peak wavelength of the red LED (R) is 625 nm, and the full width at half maximum is 15 nm; the peak wavelength of the yellow LED (Y) is 595 nm, and the full width at half maximum is 20 nm; the peak wavelength of the green LED (G) is 520 nm, and the full width at half maximum is 40 nm; the peak wavelength of the blue LED (B) is 470 nm, and the full width at half maximum is 30 nm; the peak wavelength of the cyan LED (C) is 460 nm, and the full width at half maximum is 25 nm; and the white LED (W) has peak wavelengths of 465 nm and 600 nm;

[0065] S2. Data collection and processing: Collect the luminous flux, color temperature, color rendering index, and the circadian action factor (CAF) of different age groups of the light source A data, and process the collected data so that the data is normalized within the range of [0, 1];

[0066] S3. BP neural network design: It includes: an input layer 101 with 5 nodes, corresponding to luminous flux, color temperature, color rendering index, rhythm stimulation factor, and user age respectively; a hidden layer 102 with 20 basis points; an output layer 103 with 6 nodes, corresponding to the duty ratios of red, yellow, green, blue, cyan, and white light sources respectively; the input layer 101 and the output layer 103 use activation functions to process the node data; the input layer 101 can receive specific parameters, and the hidden layer 102 extracts the abstract features of the data through weighted summation of weights and biases and the processing of activation functions, and the output layer 103 converts the features extracted by the hidden layer 102 into the final prediction result. This clear structural design makes the calculation process of the model in an orderly manner and is also convenient for subsequent parameter optimization and model adjustment; the mathematical expressions of the activation functions used by the input layer 101 and the output layer 103 are: When the activation function processes the data collected in S2, it needs to be initialized with a small random number to break symmetry and avoid redundancy of neural network weights; among them, for weight initialization: use the Xavier uniform distribution, with a range of [-0.447, 0.447]; for bias initialization: set it to 0 or a small random value close to 0.

[0067] S4. Neural network training model: It includes the following steps:

[0068] S41. Collect the dataset samples in S2;

[0069] S42. Divide the collected data set into a training set, a validation set, and a test set. Dividing the collected data points into a training set, a validation set, and a test set can make full use of the data resources. The validation set plays a monitoring role during the model training process, and the test set is used to finally evaluate the performance of the model;

[0070] S43. Apply weights and biases between the input layer 101, the hidden layer 102, and the output layer 103. The weight matrix from the input layer 101 to the hidden layer 102 is W1 ∈ R 20×5 , and the bias vector b1 ∈ R 20 ;

[0071] The weighted input of the hidden layer 102 is z1 = W1 × x + b1, and the activation output of the hidden layer 102 is a1 = σ(z1);

[0072] The weight matrix from the hidden layer 102 to the output layer 103 is W2 ∈ R 6×20 , and the bias vector b2 ∈ R 6 ;

[0073] The weighted input of the output layer 103 is z2 = W2 × a1 + b2, and the activation output of the output layer 103 is a2 = σ(z2);

[0074] S44. Introduce a formula for reducing errors: where e represents the magnitude of the error, n is the number of training times, y k ∈ R 6 is the predicted value of the output layer 103, is the true value of the output layer 103. Introducing the error formula can quantify the accuracy of the model prediction. During the model training process, by continuously adjusting the weights and biases, this error value is gradually reduced, and the existence of the error formula can also more intuitively evaluate and compare the model. This method can collect data on the luminous flux, color temperature, color rendering index of the light source, and the circadian photoreceptor factor CAF A of different age groups, and input the collected data into the neural network model. By training the neural network model, accurately predict and control the duty cycle of the selected six-color LED light source to achieve specific light source parameter adjustment.

[0075] In this embodiment, S2 also introduces a statistical analysis technique for identifying and removing outliers from the collected data.

[0076] Specifically, the IQR method can be used in S2 to identify and remove outliers from the data, thereby reducing the interference of noise on model training.

[0077] Preferably, the data collected in S2 may still have some missing values, which is not conducive to the training process of the neural network model. At this time, the mean, median or interpolation method can be used to supplement it to ensure the integrity and accuracy of the data set.

[0078] In this embodiment, the collection of the photobiological rhythm factors collected in S2 includes the following steps:

[0079] S21. Calculate the spectral transmittance of human eyes of different age groups. The following formula may be used to calculate the spectral transmittance:

[0080] τ(λ,A)=10 -D(λ,A) ;

[0081] Where λ is the wavelength and A is the user's age.

[0082] S22. During the collection process, a correction factor M(A) is introduced to adjust the pupil diameter. The adjusted pupil diameter can be calculated using the following formula:

[0083] M(A)=[1-c·(A-25)] 2 ;

[0084] Where c is the adjustment coefficient for adjusting the through-hole diameters of different ages in the formula described in S22, and c=0.00559

[0085] S23, the spectral power distribution P(λ) of the light source, the spectral transmittance τ(λ, A) and the correction factor M(A), as well as the area of ​​the illuminated area on the retina Aret are calculated using the non-visual spectral response curve C(λ) and the photopic spectral response curve V(λ), as shown in the attached figure. Figure 1 As shown, the light biorhythm factor CAF based on different ages was calculated A , based on the light biorhythm factor CAF at different ages A The calculation formula is as follows:

[0086]

[0087] By calculating the spectral transmittance and pupil diameter of human eyes at different ages, combined with the spectral power distribution, spectral transmittance, correction factor, retinal illumination area, non-visual spectral response curve and photopic spectral response curve, the photobiological rhythm factor CAF based on different ages is accurately calculated. A , understanding the needs and responses of people of different ages to light can provide a scientific basis for the design and improvement of lighting products.

[0088] Specifically, D(λ,A) mentioned in S21 can be expressed as:

[0089]

[0090] In the calculation process of adjusting the pupil diameter in S22, based on the pupil diameter of a 25-year-old user, the pupil diameters of different ages are adjusted by the coefficient c = 0.00559.

[0091] Preferably, the area A of the illuminated area on the retina in S23 ret is a circular area equal to the fovea and with a diameter of 1.5 mm.

[0092] In this embodiment, the input data in S4 is the correlated color temperature CCT, luminous flux Φ, color rendering index Ra, and photobiological rhythm factor CAF A and age A;

[0093] The output result in S4 is the duty ratios of the red, yellow, green, blue, cyan, and white light sources, and the duty ratios are respectively denoted as D R , D Y , D G , D B , D C , D WW .

[0094] Specifically, during the process of training the neural network model, the duty ratios D R , D Y , D G , D B , D C , D WW of the six-color LED light source in the input data can be randomly set, the color temperature is set in the range of 2700K to 5000K, the luminous flux is taken between 150lm and 500lm, and the selected training data set sample is 10,000 groups of data, and the data set sample is randomly divided into a training set, a validation set, and a test set.

[0095] Preferably, the training set accounts for 80% of all the training data samples, the validation set accounts for 10% of all the training data samples, and the test set accounts for 10% of all the training data samples.

[0096] In this embodiment, an intelligent lighting device is provided. The intelligent lighting device uses the above-mentioned light source mixing control method. The intelligent lighting device includes a circuit board 201. A light source module 202 is provided on the circuit board 201. The light source module 202 is electrically connected to a battery 206. The circuit board 201 is also provided with a receiving module 203 for receiving user input information, a control module 204 for controlling the light source module 202, and a wireless signal transmission module 205. The light source module 202, the receiving module 203, the control module 204, the wireless signal transmission module 205, and the battery 206 are all electrically connected.

[0097] Specifically, the light source module 202 includes several LEDs that can emit red, yellow, green, blue, cyan, and white light sources. Among them, the LEDs that emit white light sources are arranged at the center of the light source module 202, and the LEDs that emit red, yellow, green, blue, and cyan light sources are distributed around the LEDs that emit white light sources, and all the LEDs are electrically connected to each other;

[0098] The receiving module 203 uses an STM32 series microcontroller as the core processing unit, which is responsible for receiving the input information of the user and intelligently adjusting the lighting parameters of the LED light source according to this information;

[0099] The control module 204 uses a PT4115 component as the core part. The control module 204 receives the control instructions from the receiving module 203. When it receives the PWM signal sent by the microcontroller on the receiving module 203, the control module 204 will adjust the current of the LED light source, so as to achieve precise control of the light source brightness and color temperature;

[0100] The wireless signal transmission module 205 can receive the wireless instructions of the user and transmit them to the main control chip on the circuit board 201.

[0101] Preferably, the power supply module can provide a working voltage of 3.6V for the main control chip on the circuit board 201, and also provide a working voltage for other modules;

[0102] There are 3 LEDs that emit white light sources at the center of the light source module 202, and 2 LEDs each for the remaining red, yellow, green, blue, and cyan light sources, and they are circularly distributed around the LEDs that emit white light sources.

[0103] In this embodiment, an indoor lighting system is provided. The indoor lighting system includes the above intelligent lighting device. The indoor lighting system includes a children's room lighting system and a study lighting system. The user can input the user's age and room area parameters to the intelligent lighting device, so that the intelligent lighting device can calculate and adjust the luminous flux Φ according to the parameters input by the user. The children's room lighting system and the study lighting system provide customized lighting solutions for users of different ages, consider the eye transmittance and pupil correction factor M value of the user, adjust the color temperature and luminous flux, adapt to different learning and reading activities, and dynamically adjust the lighting parameters to adapt to the biological rhythm of the occupants.

[0104] Specifically, the children's room lighting system provides high-color-temperature and high-color-rendering-index lighting during the day and low-color-temperature and low-light biological rhythm factor lighting at night to meet the lighting needs of children at different times;

[0105] The study lighting system provides customized lighting solutions for users of different ages, takes into account the eye transmittance of the users and the pupil correction factor M value, adjusts the color temperature and luminous flux, adapts to different learning and reading activities, and dynamically adjusts the lighting parameters to adapt to the biological rhythm of the occupants.

[0106] Preferably, in the daytime lighting mode of the children's room lighting system, the correlated color temperature CCT of the LED lamp beads of the intelligent lighting device is maintained at the level of 5000K, the color rendering index Ra is not less than 92, and the ambient light biological rhythm factor CAF A is maintained at the level of 0.9, and the luminous flux Φ is maintained at the level of 500lx. In the night lighting mode, the correlated color temperature CCT of the LED lamp beads of the intelligent lighting device is in the range of 2700K to 3000K, the color rendering index Ra is not less than 90, and the photobiological rhythm factor CAF A is less than 0.3;

[0107] In the youth mode of the study lighting system, the correlated color temperature CCT is maintained at the level of 5000K. In the adult lighting mode, the correlated color temperature CCT is in the range of 3000K to 5000K. In the elderly lighting mode, the correlated color temperature CCT is maintained at the level of 4000K.

[0108] The usage method of this embodiment is as follows: The user inputs specific parameters such as the age of the user and the size of the room area on the mobile terminal. These specific parameters are transmitted to the main control chip of the circuit board 201 on the intelligent lighting device through the wireless signal transmission module 205. The main control chip transmits the signal to the receiving module 203. The microcontroller on the receiving module 203 converts the information input by the user into a control instruction for regulating the lighting parameters of the LED light source and sends it to the control module 204. The control module 204 adjusts the current of the LED light source, thereby realizing the precise control of the light source and color temperature, and making the light source adapt to the usage scenario and biological rhythm of the user.

[0109] The above embodiments are only used to illustrate the technical concept and features of the present invention, and their purpose is to enable those skilled in the art to understand the content of the present invention and implement it. It should not be used to limit the protection scope of the present invention. Any equivalent changes or modifications made according to the spirit and essence of the present invention should be covered within the protection scope of the present invention.

Claims

1. A light source mixing control method, characterized in that: The steps include: S1, light source selection, select red, yellow, green, blue, cyan and white as light sources; S2. Data collection and processing: collecting the luminous flux, color temperature, color rendering index of the light source and the photobiological rhythm factor CAF of different age groups A and processing the collected data so that the data is standardized within the interval [0,1]; S3. BP neural network design: including: The input layer has five nodes, corresponding to luminous flux, color temperature, color rendering index, rhythmic stimulation factor and user age; Hidden layer, set to 20 basis points; The output layer has 6 nodes corresponding to the duty cycles of the red, yellow, green, blue, cyan and white light sources respectively; The input layer and the output layer process the node data using activation functions; The mathematical expression of the activation function is: S4. Training the neural network model: including the following steps: S41, collecting samples of the data set described in S2; S42, dividing the data set collected in S41 into a training set, a validation set and a test set; S43, applying weights and biases between the input layer, the hidden layer and the output layer, the weight matrix from the input layer to the hidden layer is W1∈R 20×5 , bias vector b1∈R 20 ; The weighted input of the hidden layer is z1=W1×x+b1, and the activation output of the hidden layer is a1=σ(z1); The weight matrix from the hidden layer to the output layer is W2∈R 6×20 , bias vector b2∈R 6 ; The weighted input of the output layer is z2=W2×a1+b2, and the activation output of the output layer is a2=σ(z2); S44. Introduce the formula to reduce the error: Where e represents the size of the error, n is the number of training times, and y k ∈R 6 is the predicted value of the output layer, is the true value of the output layer.

2. A light source mixing control method as claimed in claim 1, characterized in that: S2 also introduced the IQR method to identify and eliminate outliers in the collected data.

3. A light source mixing control method as claimed in claim 1, characterized in that: The collection of the photobiorhythm factors collected in S2 includes the following steps: S21. Calculate the spectral transmittance of human eyes of different age groups. The following formula may be used to calculate the spectral transmittance: τ(λ,A)=10 -D(λ,A) ; Where λ is the wavelength and A is the age S22. During the collection process, a correction factor M(A) is introduced to adjust the pupil diameter. The adjusted pupil diameter can be calculated using the following formula: M(A)=[1-c·(A-25)] 2 ; Where c is the adjustment coefficient for adjusting the through-hole diameters of different ages in the formula described in S22, and c=0.00559 S23, the spectral power distribution P(λ) of the light source, the spectral transmittance τ(λ, A) and the correction factor M(A), as well as the area of ​​the illuminated area on the retina Aret, are used to calculate the photobiological rhythm factor CAF based on different ages using the non-visual spectral response curve C(λ) and the photopic spectral response curve V(λ), as shown in FIG1. A , based on the light biorhythm factor CAF at different ages A The calculation formula is as follows:

4. A light source mixing control method as claimed in claim 1, characterized in that: When using the activation function to process the data collected in S2, random numbers need to be used for initialization, where weight initialization: using Xavier uniform distribution in the range of [-0.447, 0.447]; bias initialization: set to a small random value of 0 or close to 0.

5. A light source mixing control method as claimed in claim 1, characterized in that: The input data in S4 are the correlated color temperature CCT, the luminous flux Φ, the color rendering index Ra, and the photobiological rhythm factor CAF. A and age A; The output result in S4 is the duty cycle of the red, yellow, green, blue, cyan and white light sources, and each duty cycle is recorded as D R ,D Y ,D G ,D B ,D C ,D WW .

6. An intelligent lighting device, characterized in that: Using the LED light source mixing control method described in any one of 1-5, the intelligent lighting device includes a circuit board, a light source module is provided on the circuit board, and the light source module is electrically connected to a battery. The circuit board is also provided with a receiving module for receiving user input information, a control module for controlling the light source module, and a wireless signal transmission module. The light source module, the receiving module, the control module, the wireless signal transmission module and the battery are all electrically connected.

7. The intelligent lighting device according to claim 6, characterized in that: The light source module includes a plurality of LEDs that can emit red, yellow, green, blue, cyan and white light sources, wherein the LED that emits white light source is arranged at the center of the light source module, and the LEDs that emit red, yellow, green, blue and cyan light sources are distributed around the LED that emits white light source, and the LEDs are electrically connected.

8. An indoor lighting system, characterized in that: An intelligent lighting device comprising any one of claims 6-7, wherein the indoor lighting system comprises a children's room lighting system and a study room lighting system, and a user can input user age and room area parameters into the intelligent lighting device, and the intelligent lighting device calculates and adjusts the luminous flux Φ based on the input parameters, so that the luminous flux Φ is maintained in the range of 150lx to 500lx.

9. An indoor lighting system as claimed in claim 8, characterized in that: The children's room lighting system comprises: Daytime lighting mode: The intelligent lighting device adjusts the correlated color temperature CCT of the LED lamp beads to maintain at 5000K, the color rendering index Ra is not less than 92, and the photobiological rhythm factor CAF of the environment is A maintained at 0.9 level, the luminous flux Φ maintained at 500lx level; and Night lighting mode: The intelligent lighting device adjusts the correlated color temperature CCT of the LED lamp beads to be within the range of 2700K to 3000K, the color rendering index Ra is not less than 90, and the photobiological rhythm factor CAF of the environment A Less than 0.3, the luminous flux Φ is maintained at the level of 150lx.

10. An indoor lighting system as claimed in claim 8, characterized in that: The study lighting system comprises: Youth lighting mode: the correlated color temperature (CCT) remains at 5000K; Adult lighting mode: the correlated color temperature CCT is in the range of 3000K to 5000K; and Elderly lighting mode: the correlated color temperature (CCT) is maintained at 4000K.

Images