Intelligent control method of sleep cabin lighting system

Abstract

The invention discloses an intelligent control method for a sleep cabin lighting system, and the method is high in adaptability: through monitoring the physiological indexes (such as heart rate, skin conductance and the like) and subjective pressure feeling of a user in real time, lighting parameters, such as brightness, color temperature and the like, can be dynamically adjusted, so as to meet individual demands of different users. The system can adaptively cope with environment changes and user state changes, and the optimal lighting effect is ensured. Reasonable illumination adjustment is beneficial to adjustment of the physiological rhythm of the human body, and the user is promoted to enter a good sleep state. The sleep quality can be improved and the rest experience of the user can be improved by reducing interference and pressure of illumination on the user to the maximum extent. The intelligent control system can adjust lighting parameters in real time according to user requirements, and unnecessary energy consumption waste is avoided. By optimizing the lighting strategy, the overall energy consumption can be greatly reduced, and the energy-saving target is achieved. And real-time feedback and optimization: the system can collect physiological feedback data of the user in real time and perform analysis and optimization in combination with subjective feelings of the user. Through online learning and algorithm iteration, the system can continuously improve the control precision and the user satisfaction. The user does not need complex operation, and the system can automatically provide the best lighting experience according to the user state. The control burden of the user is relieved, and the usability and friendliness of the system are improved.

Description

Technical Field

[0001] The present invention belongs to the field of computer system engineering, and particularly relates to an intelligent control method for a lighting system of a sleep cabin. Background Art

[0002] Sleep is very important. If one doesn't rest well, the biological clock is easily disrupted, and it may cause sleep disorders and fatigue to a certain extent, resulting in reduced work efficiency or even health damage. For example, when a person is sleepy and wants to sleep, but the light is still on at this time. If getting up to turn off the light, the drowsiness may wake up. Or, when just waking up and not fully opening the eyes yet, turning on the light at this time will be significantly dazzling and affect eyesight health. Now, many people with sleep disorders use sleep cabins for conditioning. Therefore, the automatic control of the light inside the sleep cabin is very important. How to let users enjoy high-quality and comfortable sleep is an important issue at present. Summary of the Invention

[0003] In view of the defects existing in the above-mentioned prior art, the present invention provides an intelligent control method for a lighting system of a sleep cabin, including the following steps:

[0004] Step S101: Real-time monitor the physiological index data of the user through a near-infrared spectrum sensor, where the physiological index data includes heart rate, heart rate variability, pupil diameter, and body temperature;

[0005] Step S103: Input the preprocessed physiological index data into a pre-trained correlation model between physiological indexes and subjective stress feelings to estimate the stress level of the user in real time;

[0006] Step S105: Infer the current mental state of the user based on the stress level;

[0007] Step S107: Control the lighting brightness and color temperature of the sleep cabin based on the mental state.

[0008] Among them, the correlation model between physiological indexes and subjective stress feelings adopts a three-layer feedforward neural network model.

[0009] Among them, assuming that the physiological indexes include heart rate HR, heart rate variability HRV, pupil diameter PD, and body temperature T, and the subjective stress feelings include fatigue degree Fatigue and stress degree Stress, the input layer of the three-layer feedforward neural network model has four neurons corresponding to HR, HRV, PD, and T, the first hidden layer has 8 neurons, the second hidden layer has 12 neurons, and both the first hidden layer and the second hidden layer use the ReLU activation function. The output layer has 2 neurons corresponding to Fatigue and Stress; the loss function is L.

[0010] Among them, the association model between physiological indicators and subjective stress perception in step S103 is expressed by the following formula:

[0011] The first hidden layer is h1 = ReLU(W1 * [HR, HRV, PD, T] T + b1);

[0012] The second hidden layer is h2 = ReLU(W2 * h1 + b2);

[0013] The output layer is Fatigve = W 31 * h2 + b 31 ; Stress = W 32 * h2 + b 32 where W1, W2, W 31 , W 32 are weight matrices, and b1, b2, b 31 , b 32 are bias vectors, and ReLU(x) = max(0, x) is the ReLU activation function;

[0014] The loss function is L = (1 / N) * ∑((Fatigue pred - Fatigue actual ) 2 + (Stress pred - Stress actual ) 2 ); where N is the number of training samples, Fatigue pred , Stress pred is the predicted output of the model, and Fatigue actual , Stress actual are the actual values.

[0015] Among them, the following steps are used to optimize the parameters of the association model between physiological indicators and subjective stress perception: Calculate the gradients of the parameters of each layer through backpropagation, and use the gradient descent method to update the parameters to minimize the loss function L.

[0016] Among them, calculating the gradients of the parameters of each layer through backpropagation includes:

[0017] Assume that the predicted values of the output layer are Fatigue pred , Stress pred , calculate the loss gradient of the output layer

[0018] Use the chain rule to recursively calculate the gradients of the second hidden layer and the first hidden layer;

[0019] Obtain the gradients of all parameters.

[0020] Among them, updating the parameters using the gradient descent method to minimize the loss function MSE includes:

[0021] Using the gradient descent method to update the parameters to reduce the loss function MSE;

[0022] The update formula is η is the learning rate;

[0023] Repeat the above steps for the entire training set until the model converges.

[0024] Among them, the step S107 is controlled by the following function. Assume Brightness target = f(Fatigue pred , Sttess pred )

[0025] = 100 - 20 * Fatigue pred - 10 * Stress pred ;

[0026] ColorTemp target = g(Fatigue pred , Sttess pred )

[0027] = ColorTemp target = 6500 - 1000 * Fatigue pred - 500 * Stress pred ; Among them, Brightness target , ColorTemp target represent the target values of illumination brightness and color temperature; Brightness current , ColorTemp current represent the current values of illumination brightness and color temperature;

[0028] Brightness error = Brightness target - Brightness current ;

[0029] ColorTemp error = ColorTemp target - ColorTemp current , respectively represent the adjustment errors of illumination brightness and color temperature; then

[0030] Brightness output = K p * Brightness error + Ki *∫Brightness error +K d *d(Brightness error ) / dt;

[0031] ColorTemp output =K p *ColorTemp error +K i *∫ColorTemp error +K d *d(ColorTemp error ) / dt, representing the output value of the control function.

[0032] Among them, the final illumination brightness and color temperature are determined according to the above output value:

[0033] Brightness = max(0, min(100, Brightness current +Brightness output ));

[0034] ColorTemp = max(2000, min(6500, ColorTemp current +ColorTemp output ));

[0035] The present invention also proposes an intelligent control system for a sleep cabin lighting system, including:

[0036] A near-infrared spectrum sensor for real-time monitoring of physiological index data of a user, where the physiological index data includes heart rate, heart rate variability, pupil diameter, and body temperature;

[0037] A model input module for inputting the preprocessed physiological index data into a pre-trained correlation model of physiological indexes and subjective stress feelings to estimate the stress level of a user in real time;

[0038] A psychological state inference module for inferring the current psychological state of a user based on the stress level;

[0039] A control output module for controlling the illumination brightness and color temperature of the sleep cabin based on the psychological state.

[0040] Compared with the prior art, the present invention has the following advantages:

[0041] Strong adaptability:

[0042] By real - time monitoring of users' physiological indicators (such as heart rate, skin conductance, etc.) and subjective stress feelings, lighting parameters such as brightness, color temperature, etc. can be dynamically adjusted to meet the personalized needs of different users.

[0043] The system can adaptively respond to environmental changes and changes in user status to ensure the best lighting effect.

[0044] Improve sleep quality:

[0045] Reasonable lighting adjustment helps regulate the body's biological rhythm and promotes users to enter a good sleep state.

[0046] By minimizing the interference and stress of lighting on users, sleep quality can be improved and the rest experience of users can be enhanced.

[0047] Energy - saving effect:

[0048] The intelligent control system can adjust lighting parameters in real - time according to user needs, avoiding unnecessary energy consumption waste.

[0049] By optimizing the lighting strategy, the overall energy consumption can be significantly reduced to achieve the energy - saving goal.

[0050] Real - time feedback and optimization:

[0051] The system can collect users' physiological feedback data in real - time and analyze and optimize it in combination with users' subjective feelings.

[0052] Through online learning and algorithm iteration, the system can continuously improve control accuracy and user satisfaction.

[0053] Simple to use:

[0054] Users do not need complex operations, and the system can automatically provide the best lighting experience according to the user status.

[0055] It reduces the operation burden of users and improves the usability and friendliness of the system. Description of the Drawings

[0056] By referring to the following detailed description with reference to the drawings, the above - mentioned and other purposes, features, and advantages of the exemplary embodiments of the present disclosure will become easily understood. In the drawings, several embodiments of the present disclosure are shown in an exemplary rather than restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0057] Figure 1 It is a flowchart showing an intelligent control method for a sleep cabin lighting system according to an embodiment of the present invention. Detailed Embodiments

[0058] To make the objectives, technical solutions and advantages of the present invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings. Apparently, the described embodiments are only a part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0059] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "said" and "the" used in the embodiments of the present invention and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. "Plurality" generally includes at least two.

[0060] It should be understood that although the terms first, second, third, etc. may be used in the embodiments of the present invention to describe..., these... should not be limited to these terms. These terms are only used to distinguish.... For example, without departing from the scope of the embodiments of the present invention, the first... may also be referred to as the second..., and similarly, the second... may also be referred to as the first....

[0061] It should be understood that the term "and / or" used herein is only a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects before and after.

[0062] Depending on the context, the words "if", "when" as used herein may be interpreted as "when" or "when...", or "in response to determining" or "in response to detecting". Similarly, depending on the context, the phrase "if determined" or "if detected (stated condition or event)" may be interpreted as "when determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)".

[0063] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a commodity or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such commodity or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the commodity or device including the said element.

[0064] The optional embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0065] Embodiment 1

[0066] As shown in Figure 1 the figure, the present invention discloses an intelligent control evaluation method for a sleep cabin lighting system, including the following steps:

[0067] Step S101: Real-time monitor the physiological index data of the user through a near-infrared spectrum sensor, where the physiological index data includes heart rate, heart rate variability, pupil diameter, and body temperature;

[0068] Step S103: Input the preprocessed physiological index data into a pre-trained correlation model between physiological indexes and subjective stress feelings to estimate the user's stress level in real time;

[0069] Step S105: Infer the user's current mental state based on the stress level;

[0070] Step S107: Control the lighting brightness and color temperature of the sleep cabin based on the mental state.

[0071] Embodiment 2

[0072] An intelligent control method for a sleep cabin lighting system proposed by the present invention includes the following steps:

[0073] Step S101: Real-time monitor the physiological index data of the user through a near-infrared spectrum sensor, where the physiological index data includes heart rate, heart rate variability, pupil diameter, and body temperature;

[0074] Step S103: Input the preprocessed physiological index data into a pre-trained correlation model between physiological indexes and subjective stress feelings to estimate the user's stress level in real time;

[0075] Step S105: Infer the user's current mental state based on the stress level;

[0076] Step S107: Control the lighting brightness and color temperature of the sleep cabin based on the mental state.

[0077] Among them, the correlation model between physiological indexes and subjective stress feelings adopts a three-layer feedforward neural network model.

[0078] Among them, assuming that the physiological indexes include heart rate HR, heart rate variability HRV, pupil diameter PD, and body temperature T, and the subjective stress feelings include fatigue degree Fatigue and stress degree Stress, the input layer of the three-layer feedforward neural network model has four neurons corresponding to HR, HRV, PD, and T, the first hidden layer has 8 neurons, the second hidden layer has 12 neurons, and both the first hidden layer and the second hidden layer use the ReLU activation function, and the output layer has 2 neurons corresponding to Fatigue and Stress; the loss function is L.

[0079] Among them, the association model between physiological indicators and subjective stress perception in step S103 is expressed by the following formula:

[0080] The first hidden layer is h1 = ReLU(W1 * [HR, HRV, PD, T] T + b1);

[0081] The second hidden layer is h2 = ReLU(W2 * h1 + b2);

[0082] The output layer is Fatigue = W 31 * h2 + b 31 ; Stress = W 32 * h2 + b 32 where W1, W2, W 31 , W 32 are weight matrices, and b1, b2, b 31 , b 32 are bias vectors, and ReLU(x) = max(0, x) is the ReLU activation function;

[0083] The loss function is L = (1 / N) * ∑((Fatigue pred - Fatigue actual ) 2 + (Stress pred - Stress actual ) 2 ); where N is the number of training samples, Fatigue pred , Stress pred are the predicted outputs of the model, and Fatigue actual , Stress actual are the actual values.

[0084] Among them, the following steps are used to optimize the parameters of the association model between physiological indicators and subjective stress perception: Calculate the gradients of the parameters of each layer through backpropagation, and use the gradient descent method to update the parameters to minimize the loss function L.

[0085] Among them, calculating the gradients of the parameters of each layer through backpropagation includes:

[0086] Assume that the predicted values of the output layer are Fatigue pred , Stress pred , and calculate the loss gradient of the output layer

[0087] Use the chain rule to recursively calculate the gradients of the second hidden layer and the first hidden layer;

[0088] Obtain the gradients of all parameters.

[0089] Using the chain rule, recursively calculate the gradients of the second hidden layer and the first hidden layer, specifically including:

[0090] Gradient of the second hidden layer:

[0091] The output of the second hidden layer is h (2) = f(w (2) h (1) + b (2) ), where f(·) is the activation function. According to the chain rule, for L with respect to W (2) and b (2) , we can obtain:

[0092]

[0093] Substituting the above gradient formula, we get:

[0094]

[0095]

[0096] Gradient of the first hidden layer:

[0097] The output of the first hidden layer is h (1) = f(W (1) x + b (1) ), where x is the input; according to the chain rule, for L with respect to W (1) and b (1) taking the derivative, we can obtain:

[0098]

[0099] Substituting the above gradient formula, we get:

[0100]

[0101] where the superscript (1) represents the first hidden layer and the superscript (2) represents the second hidden layer.

[0102] Through the above steps, the gradients of all parameters (W1, b1, W2, b2, W3, b3) can be calculated. These gradients can be used to update the parameters, thereby training a better neural network model.

[0103] Among them, using the gradient descent method to update the parameters to minimize the loss function MSE includes:

[0104] Using the gradient descent method to update the parameters to reduce the loss function MSE;

[0105] The update formula is η is the learning rate;

[0106] Repeat the above steps for the entire training set until the model converges.

[0107] Among them, the step S107 is controlled by the following function. Assume that Brightness target = f(Fatigue pred , Stress pred )

[0108] = 100 - 20 * Fatigue pred - 10 * Stress pred

[0109] ColorTemp target = g(Fatigue pred , Stress pred )

[0110] = ColorTemp target = 6500 - 1000 * Fatigue pred - 500 * Stress pred ; where Brightness target , ColorTemp target represent the target values of illumination brightness and color temperature; Brightness current , ColorTemp current represent the current values of illumination brightness and color temperature;

[0111] Brightness error = Brightness target - Brightness current ;

[0112] ColorTemp error = ColorTemp target - ColorTemp current , respectively representing the adjustment errors of illumination brightness and color temperature; then

[0113] Brightness output = K p * Brightness error + K i * ∫Brightness error + K d * d(Brightness error ) / dt; ColorTemp output = K p * ColorTemp error + K i*∫ColorTemp error +K d *d(ColorTemp error ) / dt, which represents the output value of the control function.

[0114] Among them, the final illumination brightness and color temperature are determined according to the above output value:

[0115] Brightness = max(0, min(100, Brightness eurrent +Brightness output ));

[0116] ColorTemp = max(2000, min(6500, ColorTemp current +ColorTemp output ));

[0117] In order to better determine K p , K i , K d These three parameters are automatically adjusted by the following steps.

[0118] Assume that the dynamic characteristics of the lighting system can be represented by a second-order transfer function model:

[0119] G(s) = K / (s 2 +2ζωns + ωn 2 ), where K is the system gain, ζ is the damping ratio, and ωn is the natural frequency.

[0120] According to the performance indicators of the lighting system, such as fast response, small overshoot, good stability, etc., assume an ideal second-order closed-loop pole distribution:

[0121] s = -ζωn ± jωn√(1 - ζ 2 ); Assume the desired damping ratio ζ = 0.707 and the natural frequency ωn = 5 rad / s.

[0122] Substitute the closed-loop poles into the characteristic equation of the system:

[0123] s 2 +2ζωns + ωn 2 +K p *s + K i +K d *s = 0;

[0124] Solve to obtain the initial expression of the parameters:

[0125] K p = 2ζωn; K i = ωn 2; K d = (2ζωn - K p ) / ωn。

[0126] Substituting the numerical values for calculation gives the initial parameter values: K p = 7; K i = 25; K d = 0.4。

[0127] Due to changes in environmental conditions, the system parameters K, ζ, ωn may change, and then an online optimization method is adopted to automatically adjust the parameters.

[0128] For example, by minimizing a certain performance index (such as response time, overshoot, etc.) to optimize the closed-loop pole position, and updating the parameters accordingly.

[0129] For example, within each sampling period, we can use the system state (Brightness error , ColorTemp error ) of the previous sampling period to calculate the current gradient information, and adjust the values of K p , K i , K d thereby continuously improving the system performance.

[0130] Apply the optimized K p , K i , K d parameters to the system, observe the responses of illumination brightness and color temperature, and ensure that the expected performance indicators are met.

[0131] Example 3

[0132] The present invention also proposes an intelligent control system for a sleep cabin lighting system, which includes:

[0133] A near-infrared spectrum sensor for real-time monitoring of the physiological index data of the user, and the physiological index data includes heart rate, heart rate variability, pupil diameter, and body temperature;

[0134] A model input module for inputting the preprocessed physiological index data into a pre-trained correlation model of physiological indexes and subjective stress feelings to estimate the stress level of the user in real time;

[0135] A psychological state inference module for inferring the current psychological state of the user based on the stress level;

[0136] A control output module for controlling the illumination brightness and color temperature of the sleep cabin based on the psychological state.

[0137] Example 4

[0138] Embodiments of the present disclosure provide a non-volatile computer storage medium storing computer-executable instructions that can execute the method steps described in the above embodiments.

[0139] It should be noted that the computer-readable medium in the present disclosure may be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0140] The above computer-readable medium may be included in the above electronic device; or may exist separately without being assembled into the electronic device.

[0141] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or it may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0142] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a portion of code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based system that performs the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0143] The units described in the embodiments of the present disclosure may be implemented in software or in hardware. In some cases, the name of the unit does not constitute a limitation on the unit itself.

[0144] The preferred embodiments of the present invention are described above to make the spirit of the present invention clearer and easier to understand, and are not intended to limit the present invention. Any modifications, substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection defined by the appended claims of the present invention.

Claims

1. An intelligent control method for a sleeping cabin lighting system, characterized in that: The following steps are involved: Step S101, monitoring the user's physiological index data in real time through a near infrared spectrum sensor, wherein the physiological index data includes heart rate, heart rate variability, pupil diameter and body temperature; Step S103, inputting the preprocessed physiological index data into the pre-trained physiological index and subjective stress perception association model to estimate the user's stress level in real time; Step S105: inferring the user's current psychological state based on the stress level; Step S107: Control the lighting brightness and color temperature of the sleeping cabin based on the psychological state.

2. The method according to claim 1, characterized in that: The physiological index and subjective stress perception association model adopts a three-layer feedforward neural network model.

3. The method according to claim 2, characterized in that: Assume that physiological indicators include heart rate HR, heart rate variability HRV, pupil diameter PD and body temperature T, and subjective stress perception includes fatigue Fatigue and stress Stress. The input layer of the three-layer feedforward neural network model is four neurons, corresponding to HR, HRV, PD, and T. The hidden layer one is 8 neurons, and the hidden layer two is 12 neurons. Both the hidden layer one and the hidden layer two use the ReLU activation function. The output layer is 2 neurons, corresponding to Fatigue and Stress. The loss function is L.

4. The method according to claim 3, characterized in that: The physiological index and subjective stress perception correlation model in step S103 is expressed by the following formula: Hidden layer 1 is h1=ReLU(W1*[HR,HRV,PD,T] T +b1); The second hidden layer is h2=ReLU(W2*h1+b2); The output layer is Fatigue = W 31 *h2+b 31 ; Stress = W 32 *h2+b 32 , where W1, W2, W 31 , W 32 is the weight matrix, b1, b2, b 31 , b 32 is the bias vector, BeLU(x)=max(0,x) is the ReLU activation function; The loss function is L = (1 / N)*Fatigue pred -Fatigue actual ) 2 +(Sttess pred -Stress actual ) 2 ), where N is the number of training samples, Fatigue pred , Stress pred is the model prediction output, Fatigue actual ,Sttess actual is the actual value.

5. The method according to claim 4, characterized in that: The following steps are used to optimize the parameters of the model associating the physiological index with the subjective pressure perception: the gradient of the parameters of each layer is calculated by back propagation, and the parameters are updated by the gradient descent method to minimize the loss function L.

6. The method according to claim 5, characterized in that The calculation of the gradient of each layer parameter by back propagation includes: Assume that the predicted value of the output layer is Fatigue pred , Stress pred , calculate the loss gradient of the output layer Using the chain rule, recursively calculate the gradients of hidden layer 2 and hidden layer 1; Get the gradients of all parameters.

7. The method according to claim 5, characterized in that The method of updating parameters by gradient descent method to minimize the loss function MSE includes: Use gradient descent method to update parameters to reduce the loss function MSE; The update formula is η is the learning rate; Repeat the above steps for the entire training set until the model converges.

8. The method according to claim 1, characterized in that: The step S107 is controlled by the following function, assuming that Brightness target =f(Fatigue pred , Stress pred )=100-20*Fatigue pred -10*Stress pred ; ColorTemp target =g(Fatigue pred ,Stress pred )=ColorTemp target =6500-1000*Fatigue pred -500*Stress pred ; Among them, Brightness target , ColorTemp target Indicates the target value of lighting brightness and color temperature; Brightness eurrent , ColorTemp current Indicates the current value of lighting brightness and color temperature; Brightness error =Brightness target -Brightness current ; ColorTemp error =ColorTemp target -ColorTemp current , respectively represent the adjustment errors of lighting brightness and color temperature; then Brightness outpui =K p *Brightness erro r+K i *∫Brightness error +K d *d(Brightness error ) / dt; ColorTemp output =K p *ColorTemp error +K i *∫ColorTemp error +K d *d(ColorTemp error ) / dt, which represents the output value of the control function.

9. The method according to claim 8, characterized in that Determine the final lighting brightness and color temperature based on the above output values: Brightness=max(0,min(100,Brightness eurrent +Brightness output )); ColorTemp=max(2000,min(6500,ColorTemp current +ColorTemp output ))。 10. An intelligent control system for a sleeping cabin lighting system, comprising: A near-infrared spectroscopy sensor, which is used to monitor the user's physiological index data in real time, the physiological index data including heart rate, heart rate variability, pupil diameter and body temperature; A model input module, which is used to input the pre-processed physiological index data into a pre-trained physiological index and subjective stress perception association model to estimate the user's stress level in real time; A psychological state inference module, which is used to infer the user's current psychological state based on the stress level; A control output module is used to control the lighting brightness and color temperature of the sleeping cabin based on the psychological state.

Images