LED spectrum parameter weighted analysis method

By performing Z-score standardization and Fourier transform on the LED spectrum signal, frequency domain features are extracted and weighted averaged, the problem of failure to effectively analyze the impact of LED spectrum on the human body in the existing technology is solved, and the field of human-caused lighting is realized. Scientific analysis and lighting design optimization.

CN119046619APending Publication Date: 2024-11-29NANCHANG KERIP OPTOELECTRONICS TECH CO LTD +1

Patent Information

Application Number
CN202411382985.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-30
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

The prior art has failed to effectively process LED spectral parameters to intuitively understand their impact on the human body, resulting in a lack of scientific analysis and optimization of the field of human-caused lighting.

Method used

Using the LED spectral parameter weighting analysis method, the LED spectral signal is preprocessed and feature extracted through Z-score normalization and Fourier transform, eliminates the dimension difference, converts the time domain signal into the frequency domain signal, extracts the main frequency components, and The impact factor is calculated by weighted average.

Benefits of technology

The accurate processing, reconstruction and integration of LED spectral signals is achieved, and practical analysis results for human-causing lighting are provided, ensuring the accuracy and reliability of the analysis results, and effectively assessing the impact of LED spectra on human health.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119046619A_ABST
    Figure CN119046619A_ABST
Patent Text Reader

Abstract

The invention discloses an LED spectrum parameter weighted analysis method, which can obtain an influence factor of a spectrum on a human body through an LED spectrum parameter so as to evaluate the influence on the health of the human body. The method comprises the following steps: step 1, performing Z-score standardization processing on an LED spectral signal, eliminating abnormal values, and improving data accuracy; step 2, Fourier transform is utilized to convert the data processed in the step 1 from a time domain to a frequency domain, and characteristics of the LED spectral signals in the frequency domain are found out; 3, performing Fourier inverse transformation and inverse standardization on the processed data to form reduced data; and 4, carrying out weighted averaging by using the reduced data to obtain an independent influence factor. The method has the advantages that unrelated spectral signals can be effectively processed, reconstructed and integrated, and the method has practicability in exploring the influence of the LED spectral signals on the human body in the field of human factor illumination.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a method for analyzing the influence of LED spectrum on human factors, in particular to a weighted analysis method of LED spectrum parameters. Background Art

[0002] Human-centered lighting is a lighting design concept that aims to improve human health, comfort and productivity by optimizing the light environment. With the development of society, people are increasingly aware that the impact of light on human physiology and psychology is not limited to lighting and visual needs, but also includes regulating the biological clock, improving mood and promoting work efficiency. In modern society, many people work and live indoors for a long time, lacking natural light. This situation may cause biological clock disorders, affecting sleep quality and overall health.

[0003] Studies have shown that proper light can regulate the human endocrine system and help regulate biological rhythms, thereby improving sleep, mood, and general health. Data preprocessing is a crucial step in data analysis and machine learning. It aims to improve the quality of data and the accuracy of analysis results. It can solve some problems of data quality, data integration, data transformation, and data cleaning. Thus, it provides a high-quality data foundation for subsequent operations and analysis. Through effective preprocessing, the performance of the algorithm and the reliability of the analysis results can be significantly improved.

[0004] However, there is no method to process the spectral parameters emitted by LEDs to intuitively obtain their impact on the human body. To address this problem, this paper proposes a weighted analysis method for LED spectral parameters. Summary of the invention

[0005] The present invention can effectively process, reconstruct and integrate unrelated spectral signals, and is practical in exploring the influence of LED spectral signals on the human body in the field of human-induced lighting.

[0006] The technical solution adopted by the present invention is as follows: a method for weighted analysis of LED spectrum parameters, the method steps are as follows: Step S1, performing data preprocessing on the LED spectrum signal; the data preprocessing method adopts Z-score standardization; Step S2, using Fourier transform to convert the data processed in step S1 from the time domain to the frequency domain, obtain processed data, and find out the characteristics of the LED spectrum signal in the frequency domain; Step S3, performing inverse Fourier transform and denormalization on the data processed in step S2 to form restored data; Step S4, using the data restored in step S3 to perform weighted averaging to obtain a separate impact factor.

[0007] Furthermore, the specific steps of step S1 are: Step S11, the original matrix of the LED spectrum signal is shown in formula (1): (1); In the formula, A is the original matrix of LED spectrum signal, each row represents a set of spectrum data, and each column represents a spectrum feature of a set of spectrum data. is the first spectral feature of the first set of spectral data of the LED spectral signal, is the nth spectral feature of the first set of spectral data of the LED spectral signal, is the first spectral feature of the spectral data of the mth group of LED spectral signals, is the nth spectral feature of the spectral data of the mth group of LED spectral signals; is the average wavelength of the first group of LED spectrum signals, is the color temperature of the first group of LED spectrum signals; is the average wavelength of the mth group of LED spectrum signals; Step S12, performing Z-score normalization processing on the LED spectrum signal, and converting the original matrix A of the LED spectrum signal of formula (1) into the normalized matrix B of the LED spectrum signal of formula (2) through Z-score normalization; (2); Where B is the standardized matrix of LED spectral signal, It is the first spectral feature of the first set of spectral data of the LED spectral signal after standardization. is the nth spectral feature of the first set of spectral data of the LED spectral signal after standardization, is the first spectral feature of the spectral data of the mth group of LED spectral signals after normalization, is the nth spectral feature of the spectral data of the mth group of LED spectral signals after standardization; is the average wavelength of the first group of LED spectral signals after standardization, is the color temperature of the first group of LED spectrum signals after standardization; is the average wavelength of the mth group of LED spectrum signals after standardization; in, and ; In the formula, is the mean value of the nth spectral feature of the original matrix A of the LED spectral signal, represents the standard deviation of the nth spectral feature of the original matrix A of the LED spectral signal; r represents the number of rows; that is, the number of rows of the original matrix A of the LED spectral signal or the standardized matrix B of the LED spectral signal.

[0008] Furthermore, the specific steps of step S2 are: The nth spectral feature of the spectral data of the mth group of the standardized LED spectral signal Perform Fourier transform, let equal , the nth spectral feature of the spectral data of the mth group of LED spectral signals after standardization Transform from time domain signal to frequency domain signal, as shown in formula (3); (3); In the formula, is the frequency domain signal obtained after Fourier transform, s is a real number, N is the total number of signal points, is the time domain signal, is the base of the natural logarithm, which is 2.71828; i is the imaginary unit, and k is the frequency index in the discrete Fourier transform, which is an integer representing the kth frequency component in the frequency domain signal.

[0009] Furthermore, the specific steps of step S3 are: Step S31, for each standardized LED spectral signal in formula (3), the nth spectral feature of the spectral data of the mth group is Perform Fourier transform and extract eigenvalues, filter out signals with specific frequencies, and then perform inverse Fourier transform to obtain time domain signals, as shown in formula (4); (4); In the formula, is the restored data obtained after inverse Fourier transform; Step S32, the restored data obtained after inverse Fourier transformation Construct a reduction matrix C, as shown in formula (5); (5); In the formula, is the first spectral feature of the first group of spectral data of the LED spectral signal which has been standardized, extracted and Fourier transformed. is the first spectral feature of the spectral data of the mth group of LED spectral signals which has been standardized, feature extracted and then subjected to inverse Fourier transform; is the nth spectral feature of the first group of spectral data of the LED spectral signal which has been standardized, extracted and Fourier transformed, is the nth spectral feature of the spectral data of the mth group of LED spectral signals which have been standardized, feature extracted and then subjected to inverse Fourier transform; Step S33, denormalize each data in the restored matrix C, as shown in formula (6); = × + (6) ; In the formula, is the restored value after denormalization, that is, the yth data of each row in the restored matrix C after denormalization; is the nth spectral feature of the spectral data of the mth group of LED spectral signals after being standardized, feature extracted and inverse Fourier transformed, is the standard deviation, i.e., the standard deviation of the original matrix of the LED spectral signal in formula (1).

[0010] Furthermore, the specific steps of step S4 are: The weighted average of each row of the denormalized matrix C is shown in formula (7); (7); In the formula, D is the impact factor obtained, L is the total number of data, is the weight of the y-th data; Step S42, calculate the influence factor D obtained in each row and add them up to get the average value, so as to obtain the influence factor E of the LED spectrum signal. The higher the influence factor E is, the more positive the impact on human health is.

[0011] The advantages of the present invention are: (1) by standardizing the spectral signal emitted by the LED, the dimension difference between different features is eliminated. This allows the wavelength, color temperature, color rendering and other features to be compared and analyzed on the same scale, avoiding certain features from dominating the analysis due to the large original numerical range; (2) by converting the time domain signal into the frequency domain signal, the main frequency components and periodic characteristics in the spectrum are revealed, which helps to identify the hidden patterns and trends in the spectrum, and better extract the characteristic values. By assigning weights according to the importance of different spectra, the weighted average can calculate the comprehensive spectral characteristics, ensuring more accurate and reliable analysis results; (3) this method can effectively process, reconstruct and integrate unrelated spectral signals, and is practical in exploring the impact of LED spectral signals on the human body in the field of human-factor lighting. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is the average wavelength amplitude spectrum of the present invention.

[0013] Figure 2 It is the average wavelength phase spectrum of the present invention. DETAILED DESCRIPTION

[0014] The present invention works and is implemented in this way. The technical solution adopted by the present invention is as follows: a method for weighted analysis of LED spectrum parameters, the method steps are as follows: Step S1, performing data preprocessing on the LED spectrum signal; the data preprocessing method adopts Z-score standardization; Step S2, using Fourier transform to convert the data processed in step S1 from the time domain to the frequency domain, obtain processed data, and find out the characteristics of the LED spectrum signal in the frequency domain; Step S3, performing inverse Fourier transform and denormalization on the data processed in step S2 to form restored data; Step S4, using the data restored in step S3 to perform weighted averaging to obtain a separate impact factor.

[0015] Furthermore, the specific steps of step S1 are: Step S11, the original matrix of the LED spectrum signal is shown in formula (1): (1); In the formula, A is the original matrix of LED spectrum signal, each row represents a set of spectrum data, and each column represents a spectrum feature of a set of spectrum data. is the first spectral feature of the first set of spectral data of the LED spectral signal, is the nth spectral feature of the first set of spectral data of the LED spectral signal, is the first spectral feature of the spectral data of the mth group of LED spectral signals, is the nth spectral feature of the spectral data of the mth group of LED spectral signals; is the average wavelength of the first group of LED spectrum signals, is the color temperature of the first group of LED spectrum signals; is the average wavelength of the mth group of LED spectrum signals; Step S12, performing Z-score normalization processing on the LED spectrum signal, and converting the original matrix A of the LED spectrum signal of formula (1) into the normalized matrix B of the LED spectrum signal of formula (2) through Z-score normalization; (2); Where B is the standardized matrix of LED spectral signal, It is the first spectral feature of the first set of spectral data of the LED spectral signal after standardization. is the nth spectral feature of the first set of spectral data of the LED spectral signal after standardization, is the first spectral feature of the spectral data of the mth group of LED spectral signals after normalization, is the nth spectral feature of the spectral data of the mth group of LED spectral signals after standardization; is the average wavelength of the first group of LED spectral signals after standardization, is the color temperature of the first group of LED spectrum signals after standardization; is the average wavelength of the mth group of LED spectrum signals after standardization; in, and ; In the formula, is the mean value of the nth spectral feature of the original matrix A of the LED spectral signal, represents the standard deviation of the nth spectral feature of the original matrix A of the LED spectral signal; r represents the number of rows; that is, the number of rows of the original matrix A of the LED spectral signal or the standardized matrix B of the LED spectral signal.

[0016] Furthermore, the specific steps of step S2 are: The nth spectral feature of the spectral data of the mth group of the standardized LED spectral signal Perform Fourier transform, let equal , the nth spectral feature of the spectral data of the mth group of LED spectral signals after standardization Transform from time domain signal to frequency domain signal, as shown in formula (3); (3); In the formula, is the frequency domain signal obtained after Fourier transform, s is a real number, N is the total number of signal points, is the time domain signal, is the base of the natural logarithm, which is 2.71828; i is the imaginary unit, and k is the frequency index in the discrete Fourier transform, which is an integer representing the kth frequency component in the frequency domain signal.

[0017] Furthermore, the specific steps of step S3 are: Step S31, for each standardized LED spectral signal in formula (3), the nth spectral feature of the spectral data of the mth group is Perform Fourier transform and extract eigenvalues, filter out signals with specific frequencies, and then perform inverse Fourier transform to obtain time domain signals, as shown in formula (4); (4); In the formula, is the restored data obtained after inverse Fourier transform; Step S32, the restored data obtained after inverse Fourier transformation Construct a reduction matrix C, as shown in formula (5); (5); In the formula, is the first spectral feature of the first group of spectral data of the LED spectral signal which has been standardized, extracted and Fourier transformed. is the first spectral feature of the spectral data of the mth group of LED spectral signals which has been standardized, feature extracted and then subjected to inverse Fourier transform; is the nth spectral feature of the first group of spectral data of the LED spectral signal which has been standardized, extracted and Fourier transformed, is the nth spectral feature of the spectral data of the mth group of LED spectral signals which have been standardized, feature extracted and then subjected to inverse Fourier transform; Step S33, denormalize each data in the restored matrix C, as shown in formula (6); = × + (6) ; In the formula, is the restored value after denormalization, that is, the yth data of each row in the restored matrix C after denormalization; is the nth spectral feature of the spectral data of the mth group of LED spectral signals after being standardized, feature extracted and inverse Fourier transformed, is the standard deviation, i.e., the standard deviation of the original matrix of the LED spectral signal in formula (1).

[0018] Furthermore, the specific steps of step S4 are: The weighted average of each row of the denormalized matrix C is shown in formula (7); (7); In the formula, D is the impact factor obtained, L is the total number of data, is the weight of the y-th data; Step S42, calculate the influence factor D obtained in each row and add them up to get the average value, so as to obtain the influence factor E of the LED spectrum signal. The higher the influence factor E is, the more positive the impact on human health is.

[0019] When analyzing the impact of light on human health, the corresponding spectral characteristics are usually analyzed to determine the impact on the human body, so the detection of LED spectral characteristic values ​​and the processing of LED spectral characteristic value data become very important. When detecting LED spectrum, errors will occur due to factors such as ambient light, temperature changes, and measurement angles, which will cause errors in analyzing the impact of light on human health through spectral parameters.

[0020] Therefore, a weighted analysis method for LED spectral parameters is proposed in the present invention, which uses Z-score standardization to remove errors in multiple measurements, and then extracts the eigenvalues ​​of the processed parameters through Fourier transform and inverse transform, and then performs weighted averaging through inverse Fourier transform and inverse standardization to obtain the influencing factor.

[0021] The present invention first introduces in detail the principles of Z-score standardization and Fourier transform and inverse transform. The role of Z-score standardization is to eliminate the dimensional differences between features by converting the data into a standard normal distribution with a mean of zero and a standard deviation of one. This method helps to balance the influence of features and prevent numerical instability. After standardized data, the influence of each feature on the whole is more balanced. The role of Fourier transform is to convert time domain signals into frequency domain signals, which is convenient for analyzing the frequency components and energy distribution of the signal. In the frequency domain, filtering, denoising, signal compression and other processes can be performed more efficiently. Through Fourier transform, complex time domain signal processing becomes more intuitive and convenient, and frequency domain analysis provides a powerful tool for data processing.

[0022] Furthermore, the value of the impact factor D weight is mainly determined by using historical paper data and statistical analysis methods; The values ​​of can be seen in Tables 1 to 7.

[0023] Furthermore, Z-score is standardized, and the specific process is as follows: (1) Data centering, Z-score standardization adjusts the mean of the data to 0; all data points will be distributed around 0 after standardization; (2) Data scaling: Z-score standardization adjusts the standard deviation of the data to 1. By dividing by the standard deviation, the units of the data points become the units of the standard deviation, so that the distribution shape of the data remains unchanged but is scaled to a uniform range; (3) Unit independence: the standardized data does not have the dimensions of the original units, and all data points are in units of standard deviation, which facilitates comparison and analysis between data sets of different dimensions; (4) Dealing with the impact of outliers: Z-score standardization retains the outliers of the original data. The standardized outliers usually have a larger positive or negative Z-score, so that the outliers are still significant after standardization and are easier to identify and handle in certain analyses.

[0024] Further, Fourier transform, the specific process is: (1) Frequency analysis: Through Fourier transform, the spectral signal is converted from the wavelength domain to the frequency domain, the main frequency components in the spectrum are identified, and the color temperature, color rendering and other characteristics of the LED light source are clearly understood; (2) Filtering: filtering the converted frequency domain signal to remove noise or extract spectral information in a specific frequency range; (3) Feature extraction: After filtering, the main frequencies and frequency features are extracted and then the required feature data is obtained through inverse Fourier transform.

[0025] In the present invention, weighted averaging calculates the comprehensive spectral characteristics by assigning weights according to the importance of different spectral data, thereby ensuring the accuracy and reliability of the analysis results. Weighted averaging is particularly important when processing data from multiple sources or when it is necessary to integrate spectral data under different experimental conditions. For example, when evaluating the impact of different LED light sources on human health and comfort, weighted averaging can more effectively integrate various experimental results to obtain more representative spectral characteristics, thereby providing a scientific basis for optimizing lighting design.

[0026] Table 1 shows the weight relationship of the average wavelength in the LED spectrum of the present invention; Table 2 shows the weight relationship of the peak wavelength in the LED spectrum of the present invention; Table 3 shows the weight relationship of the half-peak bandwidth in the LED spectrum of the present invention; Table 4 shows the weight relationship of the green ratio in the LED spectrum of the present invention; Table 5 shows the weight relationship of the blue ratio in the LED spectrum of the present invention; Table 6 shows the weight relationship of light intensity in the LED spectrum of the present invention; Table 7 shows the weight relationship of color temperature in the LED spectrum of the present invention; Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Example

[0027] A method for weighted analysis of LED spectrum parameters, the implementation method is: First, obtain a set of LED spectrum detection data from a certain company. The LED spectrum detection data includes average wavelength, peak wavelength, half-peak bandwidth, green ratio, blue ratio, luminous flux, and color temperature:

[0028] These spectral data parameters are converted into the original matrix A of the LED spectral signal to obtain A ]; Using Z-score standardization, the original matrix A is converted into data with a mean of 0 and a standard deviation of 1, from which the standardized matrix B can be obtained as follows: ]; After obtaining the standardized matrix B, it is found that the mean of each column of the data is 0, the standard deviation is 1, and the structure of the data has not changed. Since the structure of the data has not changed, Fourier transform is continued to convert the signal from time domain signal to frequency domain signal.

[0029] By calculating the amplitude spectrum, the amplitude spectrum and phase spectrum of each column of data can be obtained, such as the amplitude spectrum of the average wavelength Figure 1 The phase spectrum Figure 2 shown.

[0030] The amplitude and phase eigenvalues ​​are extracted: [4.16333634e-17 4.16333634e-17 6.93889390e-17 0.00000000e+00 6.66133815e-16 3.30846461e-14 4.16333634e-17 2.12132034e+00 2.12132034e+00 2.12132034e+00 2.12132034e+00 2.12132034e+00 2.12132034e+00 2.12132034e+00 0.00000000e+00 0.00000000e+00 3.14159265e+00 0.00000000e+00 3.14159265e+00 0.00000000e+00 0.00000000e+00 1.49533450e+00 1.49533450e+00 1.49533450e+00 2.43805646e+00 -5.23598776e-01 1.49533450e+00 1.49533450e+00] After extracting its frequency domain features, the data is inversely Fourier transformed to obtain a new reduction matrix C, and then each row of the reduction matrix C is weighted averaged after denormalization; The weight values ​​are shown in Tables 1 to 7, so the impact factors can be obtained for:

[0031] Take the impact factor The average value is 1001.09, so the influence factor of this group of LED spectral parameters is 1001.09.

[0032] The innovation of the present invention lies in: (1) Data preprocessing: By standardizing the spectral signal emitted by the LED, the data has a mean of 0 and a standard deviation of 1, thereby eliminating the dimensional differences between different features. This allows the characteristics of wavelength, color temperature, color rendering, etc. to be compared and analyzed on the same scale, avoiding certain features from dominating the analysis due to the large original numerical range. (2) Frequency domain feature extraction: Applying Fourier transform to LED spectral data analysis, converting time domain signals into frequency domain signals, and extracting meaningful frequency domain features from them, helping to deeply understand the spectral components and their effects on human health. (3) Establishment of influencing factors: Allocating weights according to the importance of different spectra, and weighted averaging can calculate the comprehensive spectral characteristics, which can effectively process, reconstruct and integrate unrelated spectral signals.

[0033] The present invention studies data preprocessing and weighting algorithms in detail, and proposes a new method for weighted analysis of LED spectrum parameters. Through simulation experiments, it is proved that effective human factors influencing factors can be obtained. Then, a batch of LED spectrum data is processed by this method, and data standardization and feature value extraction can be achieved when multiple groups of tests have error values. This method can effectively obtain human factors influencing factors and is practical in solving the impact of LED light on human health.

Claims

1. A method for weighted analysis of LED spectrum parameters, characterized in that: The steps are as follows: Step S1, performing data preprocessing on the LED spectrum signal; the data preprocessing method adopts Z-score standardization; Step S2, using Fourier transform to convert the data processed in step S1 from the time domain to the frequency domain, obtain processed data, and find out the characteristics of the LED spectrum signal in the frequency domain; Step S3, performing inverse Fourier transform and denormalization on the data processed in step S2 to form restored data; Step S4, using the data restored in step S3 to perform weighted averaging to obtain a separate impact factor.

2. The LED spectrum parameter weighted analysis method according to claim 1, characterized in that: The specific steps of step S1 are: Step S11, the original matrix of the LED spectrum signal is shown in formula (1): (1); In the formula, A is the original matrix of LED spectrum signal, each row represents a set of spectrum data, and each column represents a spectrum feature of a set of spectrum data. is the first spectral feature of the first set of spectral data of the LED spectral signal, is the nth spectral feature of the first set of spectral data of the LED spectral signal, is the first spectral feature of the spectral data of the mth group of LED spectral signals, is the nth spectral feature of the spectral data of the mth group of LED spectral signals; Step S12, performing Z-score normalization processing on the LED spectrum signal, and converting the original matrix A of the LED spectrum signal of formula (1) into the normalized matrix B of the LED spectrum signal of formula (2) through Z-score normalization; (2); Where B is the standardized matrix of LED spectral signal, It is the first spectral feature of the first set of spectral data of the LED spectral signal after standardization. is the nth spectral feature of the first set of spectral data of the LED spectral signal after standardization, is the first spectral feature of the spectral data of the mth group of LED spectral signals after normalization, is the nth spectral feature of the spectral data of the mth group of LED spectral signals after standardization; in, and ; In the formula, is the mean value of the nth spectral feature of the original matrix A of the LED spectral signal, represents the standard deviation of the nth spectral feature of the original matrix A of the LED spectral signal; r represents the number of rows; that is, the number of rows of the original matrix A of the LED spectral signal or the standardized matrix B of the LED spectral signal.

3. A method for weighted analysis of LED spectrum parameters according to claim 2, characterized in that: The specific steps of step S2 are: The nth spectral feature of the spectral data of the mth group of the standardized LED spectral signal Perform Fourier transform, let equal , the nth spectral feature of the spectral data of the mth group of LED spectral signals after standardization Transform from time domain signal to frequency domain signal, as shown in formula (3); (3); In the formula, is the frequency domain signal obtained after Fourier transform, s is a real number, N is the total number of signal points, is the time domain signal, is the base of the natural logarithm, which is 2.71828; i is the imaginary unit, and k is the frequency index in the discrete Fourier transform, which is an integer representing the kth frequency component in the frequency domain signal.

4. The LED spectrum parameter weighted analysis method according to claim 3, characterized in that: The specific steps of step S3 are: Step S31, for each standardized LED spectral signal in formula (3), the nth spectral feature of the spectral data of the mth group is Perform Fourier transform and extract eigenvalues, filter out signals with specific frequencies, and then perform inverse Fourier transform to obtain time domain signals, as shown in formula (4); (4); In the formula, is the restored data obtained after inverse Fourier transform; Step S32, the restored data obtained after inverse Fourier transformation Construct a reduction matrix C, as shown in formula (5); (5); In the formula, is the first spectral feature of the first group of spectral data of the LED spectral signal which has been standardized, extracted and Fourier transformed. is the first spectral feature of the spectral data of the mth group of LED spectral signals which has been standardized, feature extracted and then subjected to inverse Fourier transform; is the nth spectral feature of the first group of spectral data of the LED spectral signal which has been standardized, extracted and Fourier transformed, is the nth spectral feature of the spectral data of the mth group of LED spectral signals which have been standardized, feature extracted and then subjected to inverse Fourier transform; Step S33, denormalize each data in the restored matrix C, as shown in formula (6); = × + (6) ; In the formula, is the restored value after denormalization, that is, the yth data of each row in the restored matrix C after denormalization; is the nth spectral feature of the spectral data of the mth group of LED spectral signals after being standardized, feature extracted and Fourier inverse transformed, is the standard deviation, i.e., the standard deviation of the original matrix of the LED spectral signal in formula (1).

5. The LED spectrum parameter weighted analysis method according to claim 4, characterized in that: The specific steps of step S4 are: The weighted average of each row of the denormalized matrix C is shown in formula (7); (7); In the formula, D is the impact factor obtained, L is the total number of data, is the weight of the y-th data; Step S42, calculate the influence factor D obtained for each row and add them up to get the average value, so as to obtain the influence factor E of the LED spectrum signal.

Citation Information

Patent Citations

  • Correlated color temperature and color rendering index analysis method of light source

    CN106872146A

  • Modulation pattern calculation device, light control device, modulation pattern calculation method, modulation pattern calculation program, and storage medium

    CN107533246A

  • Evaluation method based on LIBS spectral component analysis result

    CN111366573A

  • On-satellite data lossless compression method and system for Fourier transform spectrometer

    CN113092382A

  • Video-based non-contact multi-physiological-parameter monitoring system and method

    CN113693573A

Cited By

  • Optical fiber structure parameter measurement method, system, equipment, medium and product

    CN122108550A