Intelligent management method and system for VLC key device data

Abstract

The invention discloses a VLC key device data intelligent treatment method and system, and the method comprises the steps: carrying out the modeling of a VLC channel in a wide environment, taking a visual light source as a signal transmitter, taking a transmission medium as a channel, and receiving a signal through a signal receiving device; main indexes of VLC channel modeling are confirmed, and related data of the main indexes are clustered by adopting a K-means + + algorithm; wherein the regression model is adopted to filter out non-type data in the related data, and the evaluation standard model of the regression model is adopted to evaluate the filtered non-type data. Through the scheme of the invention, the VLC system in a wider environment can be obtained.

Description

Technical Field

[0001] The present invention relates to the field of VLC technology, and in particular to a method and system for intelligent management of VLC key component data. Background Art

[0002] Visible light communication (VLC) is a wireless communication technology that uses light and dark flashing signals emitted by objects such as fluorescent lamps or light-emitting diodes (LEDs) to transmit information. The frequency of visible light is between 400THz (wavelength 780nm) and 800THz (wavelength 375nm). When using ordinary fluorescent lamps, its transmission capacity is 10kbit / s, and when using LED lamps, it can reach 500Mbit / s.

[0003] According to the application requirements of visible light communication combining lighting and communication, the importance of each parameter is evaluated by weight, and the most important key parameter indicators are selected as the main test parameters supported by the test platform according to the constraints of lighting on LED (such as high photoelectric conversion efficiency, high color rendering index, and low light decay), as well as the requirements of the optical communication system on the light source (such as high optical power, narrow spectrum, high 3dB modulation bandwidth, good linearity, small equivalent series resistance, small junction capacitance, and good wavelength stability). Then, the test methods of these parameters are studied. For example, for the bandwidth test of LED devices, the bandwidth can be measured by the parameters of the network analyzer, or the bandwidth can be calculated by measuring the rise / fall time of the light pulse. This may be an ideal solution for building a basic test platform for visible light communication. In the later stage, the test parameters will be continuously increased, and the test specifications for the core parameters of visible light communication devices and modules will be established in the multiplexing environment of lighting communication. In the VLC system, the visible light source is used as a signal transmitter, the air is used as a transmission medium or channel, and the signal receiving device is used to receive the signal. Although VLC technology is mainly developed for indoor scenes, it is used in a wider range of environments, from underwater to vehicle-mounted communications, and the channel models of different environments are also different. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for intelligent management of VLC key component data to obtain a VLC system in a wider environment.

[0005] To achieve the above purpose, the technical solution provided by the present invention is:

[0006] The first aspect of the present application provides a method for intelligent management of VLC key component data, including:

[0007] Modeling of VLC channels in a wide range of environments, using visible light sources as signal transmitters, transmission media as channels, and signals received by signal receiving devices;

[0008] The main indicators of VLC channel modeling are confirmed, and the K-means++ algorithm is used to cluster the relevant data of the main indicators; the regression model is used to filter out non-type data in the relevant data, and the evaluation standard model of the regression model is used to evaluate the filtered out non-type data.

[0009] To optimize the above technical solutions, the specific measures taken also include:

[0010] The transmission medium is air; the main indicators include luminous flux, transmission power, channel path loss, receiving power, receiver noise and signal-to-noise ratio; the evaluation standard model of the regression model includes: correlation coefficient test r, determination coefficient R 2 and the residual standard deviation s.

[0011] The channel path loss is calculated as follows:

[0012]

[0013] Among them, L L is the channel path loss, m is the order of magnitude of the Lambertian radiation, A r is the receiving area, D is the distance between the signal receiving device and the signal transmitter, α is the angle between the normal line of the signal receiving device and the signal transmitter-signal receiving device line, β is the signal transmitter angle, and m' represents the calculation based on the Lambertian radiation level.

[0014] The received power is calculated by path loss, luminosity function and spectral power, and the calculation formula is:

[0015]

[0016] Among them, S R (λ)=L L S T (λ), L L is the channel path loss, S T (λ) is the spectral power distribution, R f (λ) is the spectral response of the filter;

[0017] When there are N light sources in a signal transmitter, the signal receiving device receives signals from multiple light sources at the same time. The total received power is the sum of the received powers of the N light sources. At this time, the channel received power is expressed as:

[0018]

[0019] Among them, P R (total) represents the total power received by the channel receiving signals from multiple light sources, P r(i) represents the received power of each light source channel.

[0020] The receiver noise and signal-to-noise ratio are calculated as follows:

[0021] Receiver noise includes three sources: ambient light noise caused by solar radiation from windows, doors, etc., and noise caused by other lighting sources such as incandescent and fluorescent lamps; shot noise caused by signal and ambient light in the photodetector; electrical preamplifier noise of the photodetector;

[0022] The receiver signal-to-noise ratio is calculated as:

[0023] The receiver noise caused by solar radiation and artificial lighting sources is filtered out, and the receiver signal-to-noise ratio is calculated based on the shot noise and the thermal noise of the photodetector circuit. The calculation formula is:

[0024]

[0025] Among them, σ shot represents the standard deviation of the shot noise, σ thermal represents the standard deviation of thermal noise.

[0026] The clustering method of the relevant data of the main indicators using the K-means++ algorithm is as follows:

[0027] Step 1: Randomly select a center u among the data points i ;

[0028] Step 2: For each data point x that has not been selected, calculate its distance from the selected center u i The distance between j represents the number of data points that have not yet been selected;

[0029] Step 3: Use the weighted probability distribution to randomly select a new data point as the new center, where the probability of the selected point z is the same as is proportional to;

[0030] Step 4: Repeat steps 2 and 3 until K centers are selected so that K = j;

[0031] Step 5: Perform K-means clustering.

[0032] The regression model formula is as follows:

[0033] ||Xθ-y|| 2 +||Γθ|| 2

[0034] Among them, X represents the input, y represents the predicted output result, |||| represents regularization, θ represents the fitting hyperparameter, and Γ represents the objective training result;

[0035] Let the objective training result Γ be the fitting value aI to prevent overfitting. The calculation process is as follows:

[0036] θ(a)=(X T X+aI) -1 X T y

[0037] Among them, I represents the identity matrix, t represents the weight constant, a represents the weight of the identity matrix, and θ(a) represents finding θ when a is determined.

[0038] The second aspect of the present application provides a VLC key component data intelligent management system, including:

[0039] The VLC channel modeling building block is used to model VLC channels in a wide range of environments, using a visible light source as a signal transmitter, a transmission medium as a channel, and a signal receiving device to receive the signal;

[0040] The clustering module is used to confirm the main indicators of VLC channel modeling and cluster the relevant data of the main indicators using the K-means++ algorithm;

[0041] The regression evaluation module is used to filter out non-type data in the relevant data by using the regression model, and to evaluate the filtered non-type data by using the evaluation standard model of the regression model.

[0042] The third aspect of the present application provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for intelligent data management of key VLC components as described above is implemented.

[0043] The fourth aspect of the present application provides a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute the VLC key component data intelligent management method as described above.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] Although VLC technology is mainly developed for indoor scenes, it can also be used in a wider range of environments, from underwater to vehicle-mounted communications, and the channel models of different environments are also different. The present invention models VLC channels in a wide range of environments. In the VLC system designed by the present invention, a visible light source is used as a signal transmitter, air is used as a transmission medium or channel, and a signal receiving device is used to receive signals. It mainly includes five indicators: luminous flux, transmission power, channel loss, receiving power, receiver noise and signal-to-noise ratio. The K-means++ algorithm is used to cluster the relevant data of the above five indicators.

[0046] The present invention uses a regression model to filter out unnecessary and insignificant non-type data in the data. In order to prevent the regression model from filtering out valid data or mistakenly deleting valid training data, the regression model evaluation standard model is used to evaluate the filtered data, and the relationship number test r and the determination coefficient R are used. 2 The residual standard deviation s is used as the evaluation result for users to understand the fitted curve regression from different angles. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 : Schematic diagram of the process flow of the VLC key component data intelligent management method of the present invention.

[0048] Figure 2 : Schematic diagram of the structure of the VLC key component data intelligent management system of the present invention.

[0049] Figure 3 : Flowchart of the method for clustering relevant data of main indicators. DETAILED DESCRIPTION

[0050] The above contents of the present invention are further described in detail below in the form of embodiments, but this should not be understood as the scope of the above subject matter of the present invention being limited to the following embodiments, and all technologies realized based on the above contents of the present invention belong to the scope of the present invention.

[0051] Some of the terms in the present invention are explained as follows:

[0052] Optical attenuation is generally the attenuation of a section of optical channel. Therefore, at both ends of the optical channel, one end is equipped with a light source to emit light, and the other end is equipped with an optical power meter to receive light. Light source power - measurement point light power = optical attenuation;

[0053] RF (Radio Frequency) is the abbreviation of Radio Frequency, which means the electromagnetic frequency that can be radiated into space, and the frequency range is from 300KHz to 30GHz.

[0054] Luminous flux refers to the radiation power that the human eye can perceive, which is equal to the product of the radiation energy of a certain band per unit time and the relative visibility of the band. Since the relative visibility of the human eye to light of different wavelengths is different, when the radiation power of light of different wavelengths is equal, their luminous flux is not equal.

[0055] K-means++ is an algorithm that selects initial values ​​(or "seeds") for the K-means clustering algorithm. The location of the K initial centroids has a great impact on the final clustering results and running time, so it is necessary to select K appropriate centroids.

[0056] The basic idea of ​​the K-means algorithm is still to minimize the K-means sum of squares, but in its specific implementation, a method of successive iterative correction is adopted.

[0057] In one embodiment, the present invention proposes a method for intelligent management of VLC key component data, such as Figure 1 As shown, including:

[0058] Modeling of VLC channels in a wide range of environments, using visible light sources as signal transmitters, transmission media as channels, and signals received by signal receiving devices;

[0059] The main indicators of VLC channel modeling are confirmed, and the K-means++ algorithm is used to cluster the relevant data of the main indicators; among them, the regression model is used to filter out non-type data in the relevant data, and the evaluation standard model of the regression model is used to evaluate the filtered out non-type data.

[0060] The transmission medium is air; the main indicators include luminous flux, transmission power, channel path loss, receiving power, receiver noise and signal-to-noise ratio; the evaluation standard model of the regression model includes: correlation coefficient test r, determination coefficient R 2 and the residual standard deviation s.

[0061] The spectrum integral is calculated as follows:

[0062] The luminous flux combines the luminosity function and the spectral power distribution to calculate the perceived power of the signal transmitter. The V() and S() functions are used to represent the luminosity function and the spectral power distribution respectively. The luminous flux of the signal transmitter is measured in luminous units, and the calculation formula is:

[0063]

[0064] 683 lumens / watt is the maximum luminous efficiency. Luminous efficiency is the ratio of luminous flux to radiant flux. It measures the electromagnetic energy radiated by the signal transmitter and the electrical energy required to provide visible light illumination.

[0065] The spatial integral is calculated as follows:

[0066] The luminous intensity g(0) is used to measure the brightness of the signal transmitter in a specific direction by utilizing the spatial emission characteristics of the signal transmitter. It is measured in Candela, which is the luminous flux per unit solid angle. The luminous intensity contains two important parameters: the axial intensity is defined as the luminous intensity at a solid angle of 0°, in Candela, and the half-beam angle θ. max It is the angle at which the light intensity drops to half of the axial intensity. The luminous flux is calculated using the spatial integral formula:

[0067]

[0068] The transmit power is calculated as follows:

[0069] The VLC transmitter is a light-emitting lamp. The driving circuit inputs information signals. These information signals are modulated and used as the forward modulation input current of the LED or LD. The power or intensity can be directly modulated by changing the driving current. The transmission power formula is:

[0070] P tx (mW) = I[η int E p ]

[0071] E p is the energy of each photon, in eV, I is the internal forward current, in mA, and the diode modulates the input data into a power of P in the form of current. tx A light pulse, measured in mW, travels through the channel and passes through an optical lens before being focused onto a photodetector or image sensor.

[0072] The channel path loss is calculated as follows:

[0073]

[0074] Among them, L L is the channel path loss, m is the order of magnitude of the Lambertian radiation, A r is the receiving area, D is the distance between the signal receiving device and the signal transmitter, α is the angle between the normal line of the signal receiving device and the signal transmitter-signal receiving device line, β is the signal transmitter angle, and m' represents the calculation based on the Lambertian radiation level.

[0075] The received power is calculated by path loss, luminosity function and spectral power, and the calculation formula is:

[0076]

[0077] Among them, S R (λ)=L L S T (λ), L L is the channel path loss, S T (λ) is the spectral power distribution, R f (λ) is the spectral response of the filter;

[0078] When there are N light sources in a signal transmitter, the signal receiving device receives signals from multiple light sources at the same time. The total received power is the sum of the received powers of the N light sources. At this time, the channel received power is expressed as:

[0079]

[0080] Among them, P R (total) represents the total power received by the channel receiving signals from multiple light sources, P r (i) represents the received power of each light source channel.

[0081] The receiver noise and signal-to-noise ratio are calculated as:

[0082] Receiver noise includes three sources: ambient light noise caused by solar radiation from windows, doors, etc., and noise caused by other lighting sources such as incandescent and fluorescent lamps; shot noise caused by signal and ambient light in the photodetector; electrical preamplifier noise of the photodetector;

[0083] Ambient noise from solar radiation and artificial lighting sources constitutes the noise floor, which is a DC interference. The impact of this noise can be mitigated by using an electrical high-pass filter at the receiver. Typical studies assume that the ambient noise floor remains constant in space and time.

[0084] The receiver signal-to-noise ratio is calculated as:

[0085] The receiver noise caused by solar radiation and artificial lighting sources is filtered out, and the receiver signal-to-noise ratio is calculated based on the shot noise and the thermal noise of the photodetector circuit. The calculation formula is:

[0086]

[0087] Among them, σ shot represents the standard deviation of the shot noise, σ thermal represents the standard deviation of thermal noise.

[0088] like Figure 3 As shown in the figure, the clustering method of the relevant data of the main indicators using the K-means++ algorithm is as follows:

[0089] Step 1: Randomly select a center u among the data points i ;

[0090] Step 2: For each data point x that has not been selected, calculate its distance from the selected center u i The distance between j represents the number of data points that have not yet been selected;

[0091] Step 3: Use the weighted probability distribution to randomly select a new data point as the new center, where the probability of the selected point z is the same as is proportional to;

[0092] Step 4: Repeat steps 2 and 3 until K centers are selected so that K = j;

[0093] Step 5: Perform K-means clustering.

[0094] The regression model formula is as follows:

[0095] ||Xθ-y|| 2 +||Γθ|| 2

[0096] Among them, X represents the input, y represents the predicted output result, |||| represents regularization, θ represents the fitting hyperparameter, and Γ represents the objective training result;

[0097] Let the objective training result Γ be the fitting value aI to prevent overfitting. The calculation process is as follows:

[0098] θ(a)=(X T X+aI) -1 X T y

[0099] Among them, I represents the identity matrix, t represents the weight constant, a represents the weight of the identity matrix, and θ(a) represents finding θ when a is determined.

[0100] The evaluation criteria of the regression model include the correlation coefficient test r, the determination coefficient R 2 And the residual standard deviation s:

[0101] 1) Correlation coefficient test

[0102] The correlation coefficient is a quantitative indicator that describes the closeness of the linear relationship between two variables. The value range of the correlation coefficient r is [-1, 1]. If r = 1, it indicates a completely positive correlation; if r = -1, it indicates a completely negative correlation; if r = 0, it indicates no correlation. The value of r between (0, 1) is a positive correlation, and between (-1, 0) is a negative correlation.

[0103] The calculation formula of the correlation coefficient is:

[0104]

[0105] Or written as:

[0106]

[0107] Another formula for calculating the correlation coefficient from ANOVA is:

[0108]

[0109] 2) Determination coefficient R 2

[0110] definition:

[0111]

[0112] in, Obviously R 2 ≤1.R 2 Large, indicating the observed value y i With the fitted value It is relatively close, which means that overall, the distribution of n points is closer to the curve, so R is selected. 2 Larger equations are better.

[0113] 3) Residual standard deviation s

[0114] definition:

[0115]

[0116] s can be regarded as the arithmetic root of the mean residual sum of squares, and the equation with a smaller value is better. 2 The size of gives the overall degree of fit, and s gives a value of the deviation between the observation point and the regression curve.

[0117] In another embodiment, the present invention proposes a VLC key component data intelligent management system, such as Figure 2 As shown, including:

[0118] The VLC channel modeling building block is used to model VLC channels in a wide range of environments, using a visible light source as a signal transmitter, a transmission medium as a channel, and a signal receiving device to receive the signal;

[0119] The clustering module is used to confirm the main indicators of VLC channel modeling and cluster the relevant data of the main indicators using the K-means++ algorithm;

[0120] The regression evaluation module is used to filter out non-type data in the relevant data by using the regression model, and to evaluate the filtered non-type data by using the evaluation standard model of the regression model.

[0121] In another embodiment, the present invention proposes an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the above-mentioned VLC key component data intelligent management method is implemented.

[0122] In another embodiment, the present invention provides a computer-readable storage medium storing a computer program, which enables a computer to execute the above-mentioned VLC key component data intelligent management method.

[0123] In the embodiments disclosed in the present application, the computer storage medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. The computer storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the above. More specific examples of computer storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above.

[0124] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any technician familiar with the profession, without departing from the scope of the technical solution of the present invention, according to the technical essence of the present invention, any simple modification, equivalent replacement and improvement made to the above embodiment still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for intelligent management of VLC key component data, characterized in that: include: Modeling of VLC channels in a wide range of environments, using visible light sources as signal transmitters, transmission media as channels, and signals received by signal receiving devices; The main indicators of VLC channel modeling are confirmed, and the K-means++ algorithm is used to cluster the relevant data of the main indicators; among them, the regression model is used to filter out non-type data in the relevant data, and the evaluation standard model of the regression model is used to evaluate the filtered out non-type data.

2. The VLC key component data intelligent management method according to claim 1, characterized in that: The transmission medium is air; the main indicators include luminous flux, transmission power, channel path loss, receiving power, receiver noise and signal-to-noise ratio; The evaluation standard model of the regression model includes: correlation coefficient test r, determination coefficient R 2 and the residual standard deviation s.

3. The VLC key component data intelligent management method according to claim 2, characterized in that: The channel path loss is calculated as follows: Among them, L L is the channel path loss, m is the order of magnitude of the Lambertian radiation, A r is the receiving area, D is the distance between the signal receiving device and the signal transmitter, α is the angle between the normal line of the signal receiving device and the signal transmitter-signal receiving device line, β is the signal transmitter angle, and m' represents the calculation based on the Lambertian radiation level.

4. The VLC key component data intelligent management method according to claim 2, characterized in that: The received power is calculated by path loss, luminosity function and spectral power, and the calculation formula is: Among them, S R (λ)= LL S T (λ), L L is the channel path loss, S T (λ) is the spectral power distribution, R f (λ) is the spectral response of the filter; When there are N light sources in a signal transmitter, the signal receiving device receives signals from multiple light sources at the same time. The total received power is the sum of the received powers of the N light sources. At this time, the channel received power is expressed as: Among them, P R (total) represents the total power received by the channel receiving signals from multiple light sources, P r (i) represents the received power of each light source channel.

5. The VLC key component data intelligent management method according to claim 2, characterized in that: The receiver noise and signal-to-noise ratio are calculated as follows: Receiver noise includes three sources: ambient light noise caused by solar radiation from windows, doors, etc., and noise caused by other lighting sources such as incandescent and fluorescent lamps; shot noise caused by signal and ambient light in the photodetector; electrical preamplifier noise of the photodetector; The receiver signal-to-noise ratio is calculated as: The receiver noise caused by solar radiation and artificial lighting sources is filtered out, and the receiver signal-to-noise ratio is calculated based on the shot noise and the thermal noise of the photodetector circuit. The calculation formula is: Among them, σ shot represents the standard deviation of the shot noise, σ thermal represents the standard deviation of thermal noise.

6. The VLC key component data intelligent management method according to claim 1, characterized in that: The clustering method of the relevant data of the main indicators using the K-means++ algorithm is as follows: Step 1: Randomly select a center among the data points u i; Step 2: For each data point x that has not been selected, calculate the distance between it and the selected center ui j represents the number of data points that have not yet been selected; Step 3: Use the weighted probability distribution to randomly select a new data point as the new center, where the probability of the selected point z is the same as is proportional to; Step 4: Repeat steps 2 and 3 until K centers are selected so that K = j; Step 5: Perform K-means clustering.

7. The VLC key component data intelligent management method according to claim 1, characterized in that: The regression model formula is as follows: ||Xθ-y|| 2 +||Γθ|| 2 Among them, X represents the input, y represents the predicted output result, |||| represents regularization, θ represents the fitting hyperparameter, and Γ represents the objective training result; Let the objective training result Γ be the fitting value aI to prevent overfitting. The calculation process is as follows: θ(a)=(X T X+aI) -1 X T y Among them, I represents the identity matrix, t represents the weight constant, a represents the weight of the identity matrix, and θ(a) represents finding θ when a is determined.

8. A VLC key component data intelligent management system, characterized in that: include: The VLC channel modeling building block is used to model VLC channels in a wide range of environments, using a visible light source as a signal transmitter, a transmission medium as a channel, and a signal receiving device to receive the signal; The clustering module is used to confirm the main indicators of VLC channel modeling and cluster the relevant data of the main indicators using the K-means++ algorithm; The regression evaluation module is used to filter out non-type data in the relevant data by using the regression model, and to evaluate the filtered non-type data by using the evaluation standard model of the regression model.

9. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for intelligent data management of VLC key components as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute the VLC key component data intelligent management method as described in any one of claims 1 to 7.

Images