Intelligent human factor lighting method based on multi-sensing information

Abstract

The invention discloses an intelligent human factor lighting method based on multi-sensing information, and the method comprises the following steps: collecting environment data and human physiological data, and carrying out the processing of the environment data and the human physiological data, so as to obtain the preprocessed environment data and human physiological data; an improved Kalman filtering algorithm is adopted to carry out fusion processing on the preprocessed environment data and human body physiological data, the fused data are input into an optimized BP neural network to be trained, an illumination control model is obtained, a corresponding illumination strategy is generated through the illumination control model according to the environment data and the human body physiological data, and illumination control is carried out on the environment data and the human body physiological data. And generating a control signal from the lighting strategy to drive an LED lamp module. According to the invention, the Kalman filtering algorithm is improved to process the data collected by the multi-sensor module, so that the module can adaptively cope with a noise environment with complex changes, the filtering precision is remarkably improved, and accurate data input is provided for next decision-making level data fusion.

Description

Technical Field

[0001] The present invention relates to the technical field of indoor LED intelligent lighting, and in particular to an intelligent human-factor lighting method based on multi-sensor information. Background Art

[0002] With the development of science and technology and the improvement of people's living standards, people are not only paying more and more attention to the relationship between lighting and health, but also pay more attention to the enjoyment experience of lighting. People are no longer satisfied with just having lights, but need a natural, comfortable, healthy and changeable light environment. LED lighting products have significant advantages such as high light efficiency, long life, energy saving and environmental protection. They can achieve precise dimming and color control, adapt to different atmosphere requirements, and meet different lighting scenes. In recent years, with the development of the Internet of Things and artificial intelligence technology, researchers have begun to study intelligent lighting systems, introduce wireless sensing technology and automatic control technology into lighting control systems, and control the lighting system through the information collected by sensors, making the lighting system more intelligent. However, the data collected by sensors used in most current lighting systems is not accurate enough, and they cannot adapt to complex noise environments. In addition, there are problems such as the lighting system's automatic adjustment of lighting is not fast enough and the accuracy is not high, which leaves a lot of room for improvement. At the same time, relevant studies have shown that the ambient temperature and humidity content in indoor air will have a great impact on human comfort. The illumination, color temperature, color rendering index and other parameters of ambient light are also closely related to the psychological and physiological health of the human body. Warm-colored lights are often considered to be able to affect people's emotional fluctuations and make people feel full of energy, but for those with heart and brain diseases, warm-colored lights may cause discomfort. In contrast, cold-colored lights bring a sense of calmness and calmness, which helps to soothe emotions, but for patients with depression, cold-colored lights may aggravate their discomfort. Therefore, the present invention collects information such as ambient temperature values, humidity values, human heart rate values, and blood oxygen saturation values ​​through multiple sensors, and connects them with the lighting system. The improved Kalman filter algorithm is used to first perform independent decision-making at the data level fusion, and the results after the independent decision are input into the improved BP neural network algorithm for decision-level data fusion, thereby enhancing the training efficiency and prediction accuracy of the neural network model. At the same time, the connection between multi-sensor information and light color temperature is established to achieve the purpose of adaptively adjusting the light color temperature of the lighting system according to environmental information and human physiological information, and improve the response speed of the lighting system and the accuracy of light adjustment. Summary of the invention

[0003] In view of the deficiencies in the prior art, the present invention provides an intelligent human-factor lighting method based on multi-sensor information to solve the problems raised in the above background technology.

[0004] To achieve the above object, the present invention provides the following technical solution: an intelligent human-factor lighting method based on multi-sensor information, comprising the following steps: Step S1: construct a multi-sensor module, a data fusion processing module, a terminal control module, an LED driving circuit module and an LED lamp module; Step S2: collecting environmental data and human physiological data through a multi-sensor module, and amplifying, filtering, and A / D converting the collected environmental data and human physiological data to obtain pre-processed environmental data and human physiological data; Step S3: using the improved Kalman filter algorithm in the data fusion processing module to perform data-level fusion processing on the pre-processed environmental data and human physiological data to obtain fused data, inputting the fused data into the optimized BP neural network for training, obtaining a trained lighting control model of multi-sensor information and light color temperature, and generating a corresponding lighting strategy according to the environmental data and human physiological data through the trained lighting control model of multi-sensor information and light color temperature; Step S4: generating a corresponding lighting strategy into a control signal through a terminal control module; Step S5: using the LED driving circuit module to drive the LED lamp module through the control signal. Furthermore, the improved Kalman filter algorithm performs data-level fusion processing on the preprocessed environmental data and human physiological data. The specific process is as follows: State variable definition; Set the state vector of environmental data and human physiological data It is expressed as: (1); In the formula, represents the temperature at the kth moment; represents the humidity at the kth moment; represents the heart rate at the kth moment; represents the blood oxygen saturation at the kth moment; The state vector contains key variables such as temperature, humidity, heart rate, and blood oxygen saturation. The state vector describes the system state at time step k; Equation of state: Assume that the transition of the system state at discrete time k is expressed as: (2); In the formula, Indicates that the system is at time The state vector of represents the system state transfer matrix; represents the process noise at time k-1; System state transition matrix It is expressed as: (3); Observation equation; The observation equation is assumed to be: (4); In the formula, Represents the observation value vector at the kth moment; represents the observation matrix; represents the observation noise; Observation Matrix express: (5); initialization: Before filtering begins, the initial state estimate and covariance matrix are expressed as: (6); (7); In the formula, represents the initial state vector; Indicates the initial temperature value; Indicates the initial humidity value; Indicates the initial heart rate value; Indicates the initial blood oxygen saturation value; represents the initial covariance matrix; , , , Respectively represent the initial values ​​of temperature, humidity, heart rate, and blood oxygen saturation; Prediction steps: The predicted status indicates: (8); In the formula, represents the predicted state vector; The prediction covariance matrix is ​​represented as: (9); In the formula, represents the prediction covariance matrix; Indicates The covariance matrix of the time instant; Represents the transpose of the system state transfer matrix; represents the covariance matrix of the process noise; Update steps; Calculate the Kalman gain matrix representation: (10); In the formula, represents the Kalman gain matrix; represents the transpose of the observation matrix; R represents the covariance matrix of the observation noise; The updated state estimate is represented by: (11); In the formula, represents the updated state estimate; represents the predicted state estimate; Update the covariance matrix representation: (12); In the formula, represents the updated covariance matrix; represents the identity matrix; Dynamically adjust the noise covariance: The input representation of the neural network is: (13); In the formula, represents the input vector at the kth moment; represents the measurement error; The neural network output is represented by: (14); In the formula, represents a trained neural network; Represents the noise covariance matrix of the dynamic process after the output of the neural network; represents the dynamic measurement noise covariance matrix after the output of the neural network; Training of Neural Network: The training objective function is set as: (15); In the formula, represents the objective function value; N represents the number of training samples; Represents the training sample A quantity.

[0005] Furthermore, the lighting control model of the trained multi-sensor information and light color temperature is obtained. The specific process is as follows: The BP neural network model structure is as follows: The input layer of the BP neural network: 4 neurons in the updated state estimation matrix , 4 neurons Contains temperature value, humidity value, heart rate value, blood oxygen saturation value expressed as; (16); The hidden layer of the BP neural network is expressed as: (17); 0 (18); In the formula, Represents the weighted input from the input layer to the hidden layer; Represents the weight matrix from the input layer to the hidden layer; represents the bias term of the hidden layer; represents the output of the hidden layer; represents the activation function; The output layer of the BP neural network: consists of 3 neurons, which contain warm color temperature, neutral color temperature, and cold color temperature; (19); (20); In the formula, Represents the weighted input from the hidden layer to the output layer; Represents the weight matrix from the hidden layer to the output layer; Represents the bias term of the output layer; Represents the output of the output layer; Weighted feature normalization is expressed as: (twenty one); In the formula, represents the normalized input of the i-th neuron; represents the input of the i-th neuron; represents the mean value of the i-th neuron; represents the standard deviation of the i-th neuron; represents the weighting factor of the i-th neuron; Normalized input Used as input for BP neural network training; Initialize the learning range and step size: the learning rate range is ; The current learning rate is ; Adjust the step size to ; and Respectively represent the maximum and minimum values ​​of the learning rate range; Golden section method: According to the principle of golden section method, choose the optimal adjustment step length and learning rate tuning parameters It is expressed as: (twenty two); In the formula, represents the golden ratio; Sine function adjustment: Dynamically adjust the current learning rate through the sine function , expressed as; (twenty three); In the formula, Indicates the current iteration number; t represents the maximum number of iterations; represents the sine function; The error of the output layer of the improved BP neural network is expressed as: (twenty four); In the formula, Represents the output layer error of the improved BP neural network; represents the target output; The improved BP neural network hidden layer error is expressed as: (25); In the formula, Represents the hidden layer error of the improved BP neural network; represents the derivative of the activation function; Represents the transpose of the weight matrix from the hidden layer to the output layer; Optimize the current learning rate using the golden sine subtraction optimizer To update the weights, get the updated weights and , expressed as: (26); (27); In the formula, represents the loss function; and They are the gradients of the loss function with respect to the weights; The updated weight and Back propagation is performed to the hidden layer and input layer of the BP neural network for multiple back propagation iterations to train the BP neural network, and a lighting control model of the trained multi-sensor information and light color temperature is obtained.

[0006] Furthermore, the multi-sensor module is composed of a temperature sensor, a humidity sensor, a heart rate sensor, and a blood oxygen saturation sensor; A temperature sensor and a humidity sensor are used to detect the temperature and humidity values ​​of the indoor environment as environmental data; A heart rate sensor and a blood oxygen saturation sensor are used to monitor the human heart rate value and blood oxygen saturation value as human physiological data.

[0007] Furthermore, it also includes a low-power power supply module to supply power to the multi-sensor module, the data fusion processing module, and the terminal control module.

[0008] Furthermore, the terminal control module is composed of an embedded microcontroller, a memory, a radio frequency module, a power supply circuit, a crystal oscillator circuit and a peripheral circuit; the terminal control module is used to output a control signal to drive the LED lamp module.

[0009] Compared with the existing technology, the present invention has the following beneficial effects: (1) The present invention processes the data collected by the multi-sensor module by improving the Kalman filter algorithm, so that the module can adaptively cope with complex and changing noise environments, significantly improve the filtering accuracy, and provide accurate data input for the next step of decision-making level data fusion.

[0010] (2) The present invention improves the BP neural network by combining weighted feature normalization and golden sine subtraction optimizer (GSABO) to avoid certain features from having too much influence on the training results, and searches for the best learning rate and optimization strategy to accelerate the training of the neural network, thereby better handling the feature differences and learning rate adjustment problems in the neural network and improving the performance of the BP neural network in the lighting control system.

[0011] (3) The present invention combines multiple sensor information for fusion processing, and synergistically combines multiple complementary and redundant information in space and time to obtain a consistent description of the monitored data. The collected sensor information can help people understand the environmental conditions and people's health information in real time, so as to automatically adjust the state of the lamps and lanterns to put people in the most comfortable lighting environment. This can enable people to understand their own physical conditions in a timely manner, and at the same time has a positive effect on alleviating people's psychological pressure and negative emotions. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0013] like Figure 1 As shown, the present invention provides a technical solution: an intelligent human-factor lighting method based on multi-sensor information, comprising the following steps: Step S1: construct a multi-sensor module, a data fusion processing module, a terminal control module, an LED driving circuit module and an LED lamp module; Step S2: collecting environmental data and human physiological data through a multi-sensor module, and amplifying, filtering, and A / D converting the collected environmental data and human physiological data to obtain pre-processed environmental data and human physiological data; Step S3: using the improved Kalman filter algorithm in the data fusion processing module to perform data-level fusion processing on the pre-processed environmental data and human physiological data to obtain fused data, inputting the fused data into the optimized BP neural network for training, obtaining a trained lighting control model of multi-sensor information and light color temperature, and generating a corresponding lighting strategy according to the environmental data and human physiological data through the trained lighting control model of multi-sensor information and light color temperature; Step S4: generating a corresponding lighting strategy into a control signal through a terminal control module; Step S5: using the LED driving circuit module to drive the LED lamp module through the control signal.

[0014] Among them, the improved Kalman filter algorithm performs data-level fusion processing on the pre-processed environmental data and human physiological data. The specific process is as follows: State variable definition; Set the state vector of environmental data and human physiological data It is expressed as: (1); In the formula, represents the temperature at the kth moment; represents the humidity at the kth moment; represents the heart rate at the kth moment; represents the blood oxygen saturation at the kth moment; The state vector contains key variables such as temperature, humidity, heart rate, and blood oxygen saturation. The state vector describes the system state at time step k; Equation of state: Assume that the transition of the system state at discrete time k is expressed as: (2); In the formula, Indicates that the system is at time The state vector of represents the system state transfer matrix; represents the process noise at time k-1; System state transition matrix It is expressed as: (3); Observation equation; The observation equation is assumed to be: (4); In the formula, Represents the observation value vector at the kth moment; represents the observation matrix; represents the observation noise; Observation Matrix express: (5); initialization: Before filtering begins, the initial state estimate and covariance matrix are expressed as: (6); (7); In the formula, represents the initial state vector; Indicates the initial temperature value; Indicates the initial humidity value; Indicates the initial heart rate value; Indicates the initial blood oxygen saturation value; represents the initial covariance matrix; , , , Respectively represent the initial values ​​of temperature, humidity, heart rate, and blood oxygen saturation; Prediction steps: The predicted status indicates: (8); In the formula, represents the predicted state vector; The prediction covariance matrix is ​​represented as: (9); In the formula, represents the prediction covariance matrix; Indicates The covariance matrix of the time instant; Represents the transpose of the system state transfer matrix; represents the covariance matrix of the process noise; Update steps: Calculate the Kalman gain matrix representation: (10); In the formula, represents the Kalman gain matrix; represents the transpose of the observation matrix; R represents the covariance matrix of the observation noise; The updated state estimate is represented by: (11); In the formula, represents the updated state estimate; represents the predicted state estimate; Update the covariance matrix representation: (12); In the formula, represents the updated covariance matrix; represents the identity matrix; Dynamically adjust the noise covariance: The input representation of the neural network is: (13); In the formula, represents the input vector at the kth moment; represents the measurement error; The input of the neural network is the prediction error at the current moment, the predicted state estimate, and the prediction error covariance matrix; The neural network output is represented by: (14); In the formula, represents a trained neural network; Represents the noise covariance matrix of the dynamic process after the output of the neural network; represents the dynamic measurement noise covariance matrix after the output of the neural network; Training of Neural Network: The training objective function is set as: (15); In the formula, Represents the objective function value, which measures the total error of the neural network model prediction; N represents the number of training samples; Represents the training sample A quantity.

[0015] Among them, the lighting control model of multi-sensor information and light color temperature after training is obtained. The specific process is: The BP neural network model structure is as follows: The input layer of the BP neural network: 4 neurons in the updated state estimation matrix , 4 neurons Contains temperature value, humidity value, heart rate value, blood oxygen saturation value expressed as; (16); The hidden layer of the BP neural network is expressed as: (17); (18); In the formula, Represents the weighted input from the input layer to the hidden layer; Represents the weight matrix from the input layer to the hidden layer; represents the bias term of the hidden layer; represents the output of the hidden layer; represents the activation function; The output layer of the BP neural network: consists of 3 neurons, which contain warm color temperature, neutral color temperature, and cold color temperature; (19); (20); In the formula, Represents the weighted input from the hidden layer to the output layer; Represents the weight matrix from the hidden layer to the output layer; Represents the bias term of the output layer; Represents the output of the output layer; Weighted feature normalization is expressed as: (twenty one); In the formula, represents the normalized input of the i-th neuron; represents the input of the i-th neuron; represents the mean value of the i-th neuron; represents the standard deviation of the i-th neuron; Represents the weighting factor of the i-th neuron, which is used to control the influence of each neuron; Normalized input Used as input for BP neural network training; Initialize the learning range and step size: the learning rate range is ; The current learning rate is ; Adjust the step size to ; and Respectively represent the maximum and minimum values ​​of the learning rate range; Golden section method: According to the principle of golden section method, choose the optimal adjustment step length and learning rate tuning parameters It is expressed as: (twenty two); In the formula, represents the golden ratio; Sine function adjustment: Dynamically adjust the current learning rate through the sine function , so that the current learning rate It is gradually reduced during the training process, thereby accelerating convergence, which is expressed as; (twenty three); In the formula, Indicates the current iteration number; t represents the maximum number of iterations; represents the sine function; The error of the output layer of the improved BP neural network is expressed as: (twenty four); In the formula, Represents the output layer error of the improved BP neural network; represents the target output; The improved BP neural network hidden layer error is expressed as: (25); In the formula, Represents the hidden layer error of the improved BP neural network; represents the derivative of the activation function; Represents the transpose of the weight matrix from the hidden layer to the output layer; Use the golden sine subtraction optimizer GSABO to optimize the current learning rate To update the weights, get the updated weights and , expressed as: (26); (27); In the formula, represents the loss function; and They are the gradients of the loss function with respect to the weights; The updated weight and Back propagation is performed to the hidden layer and input layer of the BP neural network to perform multiple back propagation iterations to train the BP neural network, thereby obtaining a lighting control model of the trained multi-sensor information and light color temperature; Normalizing the weighted features is to standardize or normalize the input of the i-th neuron, use the standardized or normalized input as the input of the BP neural network, and train the BP neural network so that the values ​​of different neurons have the same importance during the training process, which can avoid excessive influence of neurons on the training results of the BP neural network; then combine the golden section method to select the optimal step size and learning rate, and dynamically adjust the learning rate through the sine function, so that the learning rate is gradually reduced during the training process, thereby accelerating convergence; through this method, the BP neural network can more effectively process features such as temperature, humidity, heart rate and blood oxygen saturation, and output correct classification results (warm color temperature, neutral color temperature, cold color temperature).

[0016] Among them, the multi-sensor module consists of a temperature sensor, a humidity sensor, a heart rate sensor, and a blood oxygen saturation sensor; A temperature sensor and a humidity sensor are used to detect the temperature and humidity values ​​of the indoor environment as environmental data; A heart rate sensor and a blood oxygen saturation sensor are used to monitor the human heart rate value and blood oxygen saturation value as human physiological data.

[0017] Among them, it also includes a low-power power supply module to power the multi-sensor module, the data fusion processing module, and the terminal control module.

[0018] The terminal control module is composed of an embedded microcontroller, a memory, a radio frequency module, a power supply circuit, a crystal oscillator circuit and a peripheral circuit; the terminal control module is used to output a control signal to drive the LED lamp module.

[0019] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent human-factor lighting method based on multi-sensor information, characterized in that: The following steps are involved: Step S1: construct a multi-sensor module, a data fusion processing module, a terminal control module, an LED driving circuit module and an LED lamp module; Step S2: collecting environmental data and human physiological data through a multi-sensor module, and amplifying, filtering, and A / D converting the collected environmental data and human physiological data to obtain pre-processed environmental data and human physiological data; Step S3: using the improved Kalman filter algorithm in the data fusion processing module to perform data-level fusion processing on the pre-processed environmental data and human physiological data to obtain fused data, inputting the fused data into the optimized BP neural network for training, obtaining a trained lighting control model of multi-sensor information and light color temperature, and generating a corresponding lighting strategy according to the environmental data and human physiological data through the trained lighting control model of multi-sensor information and light color temperature; Step S4: generating a corresponding lighting strategy into a control signal through a terminal control module; Step S5: using the LED driving circuit module to drive the LED lamp module through the control signal.

2. The intelligent human-factor lighting method based on multi-sensor information according to claim 1, characterized in that: The improved Kalman filter algorithm performs data-level fusion processing on the preprocessed environmental data and human physiological data. The specific process is as follows: State variable definition; Set the state vector of environmental data and human physiological data It is expressed as: （1）； In the formula, represents the temperature at the kth moment; represents the humidity at the kth moment; represents the heart rate at the kth moment; represents the blood oxygen saturation at the kth moment; The state vector contains key variables such as temperature, humidity, heart rate, and blood oxygen saturation. The state vector describes the system state at time step k; Equation of state: Assume that the transition of the system state at discrete time k is expressed as: （2）； In the formula, Indicates that the system is at time The state vector of represents the system state transfer matrix; represents the process noise at time k-1; System state transition matrix It is expressed as: （3）； Observation equation; The observation equation is assumed to be: （4）； In the formula, Represents the observation value vector at the kth moment; represents the observation matrix; represents the observation noise; Observation Matrix express: （5）； initialization: Before filtering begins, the initial state estimate and covariance matrix are expressed as: （6）； （7）； In the formula, represents the initial state vector; Indicates the initial temperature value; Indicates the initial humidity value; Indicates the initial heart rate value; Indicates the initial blood oxygen saturation value; represents the initial covariance matrix; , , , Respectively represent the initial values ​​of temperature, humidity, heart rate, and blood oxygen saturation; Prediction steps: The predicted status indicates: （8）； In the formula, represents the predicted state vector; The prediction covariance matrix is ​​represented as: （9）； In the formula, represents the prediction covariance matrix; Indicates The covariance matrix of the moment; Represents the transpose of the system state transfer matrix; represents the covariance matrix of the process noise; Update steps; Calculate the Kalman gain matrix representation: （10）； In the formula, represents the Kalman gain matrix; represents the transpose of the observation matrix; R represents the covariance matrix of the observation noise; The updated state estimate is represented by: （11）； In the formula, represents the updated state estimate; represents the predicted state estimate; Update the covariance matrix representation: （12）； In the formula, represents the updated covariance matrix; represents the identity matrix; Dynamically adjust the noise covariance: The input representation of the neural network is: （13）； In the formula, represents the input vector at the kth moment; represents the measurement error; The neural network output is represented by: （14）； In the formula, represents a trained neural network; Represents the noise covariance matrix of the dynamic process after the output of the neural network; represents the dynamic measurement noise covariance matrix after the output of the neural network; Training of Neural Network: The training objective function is set as: （15）； In the formula, represents the objective function value; N represents the number of training samples; Represents the training sample A quantity.

3. The intelligent human-factor lighting method based on multi-sensor information according to claim 2, characterized in that: The lighting control model of the trained multi-sensor information and light color temperature is obtained. The specific process is as follows: BP neural network model structure: The input layer of the BP neural network: 4 neurons in the updated state estimation matrix , 4 neurons Contains temperature value, humidity value, heart rate value, blood oxygen saturation value expressed as; （16）； The hidden layer of the BP neural network is expressed as: （17）； （18）； In the formula, Represents the weighted input from the input layer to the hidden layer; Represents the weight matrix from the input layer to the hidden layer; represents the bias term of the hidden layer; represents the output of the hidden layer; represents the activation function; The output layer of the BP neural network: 3 neurons, 3 neurons contain warm color temperature, neutral color temperature, and cold color temperature represented as; （19）； （20）； In the formula, Represents the weighted input from the hidden layer to the output layer; Represents the weight matrix from the hidden layer to the output layer; Represents the bias term of the output layer; Represents the output of the output layer; Weighted feature normalization is expressed as: （21）； In the formula, represents the normalized input of the i-th neuron; represents the input of the i-th neuron; represents the mean value of the i-th neuron; represents the standard deviation of the i-th neuron; represents the weighting factor of the i-th neuron; Normalized input Used as input for BP neural network training; Initialize the learning range and step size: the learning rate range is ; The current learning rate is ; Adjust the step size to ; and Respectively represent the maximum and minimum values ​​of the learning rate range; Golden section method: According to the principle of golden section method, choose the optimal adjustment step length and learning rate tuning parameters It is expressed as: （22）； In the formula, represents the golden ratio; Sine function adjustment: Dynamically adjust the current learning rate through the sine function , expressed as; （23）； In the formula, Indicates the current iteration number; t represents the maximum number of iterations; represents the sine function; The error of the output layer of the improved BP neural network is expressed as: （24）； In the formula, Represents the output layer error of the improved BP neural network; represents the target output; The improved BP neural network hidden layer error is expressed as: （25）； In the formula, Represents the hidden layer error of the improved BP neural network; represents the derivative of the activation function; Represents the transpose of the weight matrix from the hidden layer to the output layer; Optimize the current learning rate using the golden sine subtraction optimizer To update the weights, get the updated weights and , expressed as: （26）； （27）； In the formula, represents the loss function; and They are the gradients of the loss function with respect to the weights; The updated weight and Back propagation is performed to the hidden layer and input layer of the BP neural network for multiple back propagation iterations to train the BP neural network, and a lighting control model of the trained multi-sensor information and light color temperature is obtained.

4. The intelligent human-factor lighting method based on multi-sensor information according to claim 3, characterized in that: The multi-sensor module consists of a temperature sensor, a humidity sensor, a heart rate sensor, and a blood oxygen saturation sensor; A temperature sensor and a humidity sensor are used to detect the temperature and humidity values ​​of the indoor environment as environmental data; A heart rate sensor and a blood oxygen saturation sensor are used to monitor the human heart rate value and blood oxygen saturation value as human physiological data.

5. The intelligent human-factor lighting method based on multi-sensor information according to claim 4, characterized in that: It also includes a low-power power supply module to supply power to the multi-sensor module, the data fusion processing module, and the terminal control module.

6. The intelligent human-factor lighting method based on multi-sensor information according to claim 5, characterized in that: The terminal control module is composed of an embedded microcontroller, a memory, a radio frequency module, a power supply circuit, a crystal oscillator circuit and a peripheral circuit; the terminal control module is used to output a control signal to drive the LED lamp module.

Images