CN115150984A

CN115150984A - LED lamp strip and control method thereof

Info

Publication number: CN115150984A
Application number: CN202210823059.5A
Authority: CN
Inventors: 黄欣贵
Original assignee: Huizhou Wisva Optoelectronics Co ltd
Current assignee: Huizhou Wisva Optoelectronics Co ltd
Priority date: 2022-07-14
Filing date: 2022-07-14
Publication date: 2022-10-04
Anticipated expiration: 2042-07-14
Also published as: CN115150984B

Abstract

The application relates to the field of intelligent control of LED lamp belts, and particularly discloses an LED lamp belt and a control method thereof. Therefore, the LED lamp strip can be adaptively adjusted according to the difference between the indoor brightness and the outdoor brightness, so that the requirements of people are met.

Description

LED lamp strip and control method thereof

Technical Field

The invention relates to the field of intelligent control of LED lamp belts, in particular to an LED lamp belt and a control method thereof.

Background

At present, with the continuous improvement of living standards of people, the requirement for beautifying the living environment is higher and higher, and in order to beautify the living environment of people, the LED lamp strip with various light source colors is arranged in a hidden groove at the top of an indoor to form a beautiful light ring, and the light ring can not only beautify the living environment, but also can bring a beautiful light for people at night.

However, the existing switch control systems of LED strips mostly adopt electronic switches controlled manually, and cannot automatically adjust the brightness of the LED strip according to the change of indoor brightness. Meanwhile, in the brightness adjustment process of the LED lamp strip, the difference between indoor brightness and outdoor brightness needs to be considered, so that the adjusted brightness of the LED lamp strip can meet the requirement of illumination, and the discomfort of human eyes caused by excessive stimulation on the human eyes can be avoided. Therefore, it is desirable to have an LED light strip that can be adaptively adjusted according to the difference between indoor brightness and outdoor brightness to meet the needs of people.

Disclosure of Invention

The present application is proposed to solve the above-mentioned technical problems. The embodiment of the application provides an LED lamp strip and a control method thereof, wherein indoor and outdoor brightness difference dynamic characteristics and illumination power dynamic correlation characteristics of the LED lamp strip on a time sequence dimension are respectively excavated through a deep learning deep neural network model, whether the illumination power of the LED lamp strip at the current time point is properly classified is judged by utilizing fusion characteristic information of the indoor and outdoor brightness difference dynamic characteristics and the illumination power dynamic correlation characteristics, and in the characteristic fusion process, the overall distribution of an illumination power time sequence characteristic vector and a brightness transfer matrix approaches to an isotropic expression space with distinction degree relatively to each other through comparison search space syntropy of characteristic values, so that the classification effect of classification characteristic vectors is enhanced. Therefore, the LED lamp strip can be adaptively adjusted according to the difference between the indoor brightness and the outdoor brightness, so that the requirements of people are met.

According to an aspect of the application, there is provided a LED strip comprising:

the illumination data and environment brightness data acquisition module is used for acquiring illumination power values, indoor brightness values and outdoor brightness values of the LED lamp belts at a plurality of preset time points including the current time point;

the environment brightness data coding module is used for respectively arranging the indoor brightness values and the outdoor brightness values of the plurality of preset time points including the current time point into input vectors and then obtaining indoor brightness characteristic vectors and outdoor brightness characteristic vectors through a first sequence coder including one-dimensional convolution layers;

the indoor and outdoor brightness data contrast coding module is used for calculating a brightness transfer matrix of the indoor brightness characteristic vector relative to the outdoor brightness characteristic vector;

the illumination data coding module is used for enabling the illumination power values of the LED lamp strip at a plurality of preset time points including the current time point to pass through a second sequence encoder comprising a one-dimensional convolution layer so as to obtain an illumination power time sequence characteristic vector;

the illumination characteristic correction module is used for correcting the characteristic value of each position in the illumination power time sequence characteristic vector to obtain a corrected illumination power time sequence characteristic vector;

the brightness characteristic correction module is used for correcting the characteristic value of each position in the brightness transfer matrix to obtain a corrected brightness transfer matrix;

the feature mapping fusion module is used for multiplying the corrected illumination power time sequence feature vector and the corrected brightness transfer matrix and mapping the high-dimensional brightness contrast change information of the corrected brightness transfer matrix to the high-dimensional feature space of the corrected illumination power time sequence feature vector to obtain a classification feature vector; and

and the illumination control result generation module is used for enabling the classified characteristic vectors to pass through a classifier to obtain a classification result, and the classification result is used for indicating that the illumination power of the LED lamp strip at the current time point should be increased or decreased.

In the above LED light strip, the ambient brightness data encoding module includes: the indoor brightness coding unit is used for arranging the indoor brightness values of the plurality of preset time points including the current time point into an indoor brightness input vector according to a time dimension; using said first sequenceThe full-connection layer of the encoder performs full-connection encoding on the indoor brightness input vector by using the following formula to extract high-dimensional implicit features of feature values of all positions in the indoor brightness input vector, wherein the formula is as follows:

wherein X is the input vector, Y is the output vector, W is the weight matrix, B is the offset vector,

represents a matrix multiplication; performing one-dimensional convolutional coding on the indoor luminance input vector by using a one-dimensional convolutional layer of the first sequence encoder according to the following formula to extract high-dimensional implicit correlation features among feature values of all positions in the indoor luminance input vector, wherein the formula is as follows:

wherein, a is the width of the convolution kernel in the X direction, F is the parameter vector of the convolution kernel, G is the local vector matrix operated with the convolution kernel function, w is the size of the convolution kernel, and X represents the input vector; the outdoor brightness coding unit is used for arranging the outdoor brightness values of the plurality of preset time points including the current time point into an outdoor brightness input vector according to the time dimension; performing full-concatenation encoding on the outdoor luminance input vector using a full-concatenation layer of the first sequence encoder to extract high-dimensional implicit features of feature values of respective positions in the outdoor luminance input vector according to the following formula:

represents a matrix multiplication; performing one-dimensional convolutional coding on the outdoor luminance input vector by using a one-dimensional convolutional layer of the first sequence encoder according to the following formula to extract high-dimensional implicit correlation features among feature values of all positions in the outdoor luminance input vector, wherein the formula is as follows:

wherein, a is the width of the convolution kernel in the X direction, F is the parameter vector of the convolution kernel, G is the local vector matrix operated with the convolution kernel function, w is the size of the convolution kernel, and X represents the input vector.

In the above LED light strip, the indoor and outdoor brightness data comparison and encoding module is further configured to: calculating the brightness transfer matrix of the indoor brightness eigenvector relative to the outdoor brightness eigenvector by the following formula;

wherein the formula is:

M＝F ₁ *F ₂

wherein F ₂ Representing the outdoor luminance feature vector, M representing the luminance transfer matrix, F ₁ Representing the indoor luminance feature vector.

In the LED light strip, the illumination data encoding module is further configured to: arranging the illumination power values of the LED lamp belts at a plurality of preset time points including the current time point into an illumination power input vector according to a time dimension; using a full-concatenation layer of the second sequence encoder to perform full-concatenation encoding on the illumination power input vector to extract high-dimensional implicit features of feature values of respective positions in the illumination power input vector, according to the following formula:

represents a matrix multiplication; performing one-dimensional convolution encoding on the illumination power input vector by using a one-dimensional convolution layer of the second sequence encoder according to the following formula to extract high-dimensional implicit correlation features among feature values of all positions in the illumination power input vector, wherein the formula is as follows:

In the above LED light strip, the lighting characteristic correction module is further configured to: correcting the characteristic value of each position in the illumination power time sequence characteristic vector according to the following formula to obtain the corrected illumination power time sequence characteristic vector;

wherein the formula is:

wherein v is _i Feature values, m, representing respective positions of the illumination power time-sequential feature vector mapped to a probability space _j,k Characteristic values, d (v), representing the respective positions of the luminance transfer matrix mapped into the probability space _i ，m _j，k ) Represents the distance between the characteristic values, and ρ is a hyperparameter.

In the above LED light strip, the brightness characteristic correction module is further configured to: correcting the characteristic value of each position in the brightness transfer matrix according to the following formula to obtain the corrected brightness transfer matrix;

wherein the formula is:

wherein v is _i Feature values, m, representing respective positions of the illumination power time-sequential feature vector mapped to a probability space _j，k Characteristic values, d (v), representing the respective positions of the luminance transfer matrix mapped into the probability space _i ，m _j,k ) Represents the distance between the characteristic values, and ρ is a hyperparameter.

In the LED lamp strip, the d (v) _i ，m _j,k ) And representing Euclidean distances between the eigenvalues of each position in the illumination power time sequence eigenvector and the eigenvalue of each position in the brightness transfer matrix.

In the LED light strip, the lighting control result generating module is further configured to: processing the classification feature vector using the classifier to obtain the classification result with the following formula: softmax { (W) _n ，B _n )：...：(W ₁ ，B ₁ ) I X }, wherein W ₁ To W _n As a weight matrix, B ₁ To B _n Is a bias vector and X is the classification feature vector.

According to another aspect of the application, a method of controlling a LED strip comprises:

the method comprises the steps of obtaining the illumination power value, the indoor brightness value and the outdoor brightness value of the LED lamp belt at a plurality of preset time points including the current time point;

respectively arranging the indoor brightness values and the outdoor brightness values of the plurality of preset time points including the current time point into input vectors, and then obtaining indoor brightness characteristic vectors and outdoor brightness characteristic vectors through a first sequence encoder including one-dimensional convolution layers;

calculating a brightness transfer matrix of the indoor brightness eigenvector relative to the outdoor brightness eigenvector;

the lighting power values of the LED lamp strip at a plurality of preset time points including the current time point are processed by a second sequence encoder comprising a one-dimensional convolution layer to obtain a lighting power time sequence characteristic vector;

correcting the characteristic value of each position in the illumination power time sequence characteristic vector to obtain a corrected illumination power time sequence characteristic vector;

correcting the characteristic value of each position in the brightness transfer matrix to obtain a corrected brightness transfer matrix;

multiplying the corrected illumination power time sequence eigenvector and the corrected brightness transfer matrix to map high-dimensional brightness contrast change information of the corrected brightness transfer matrix into a high-dimensional eigenvector space of the corrected illumination power time sequence eigenvector to obtain a classification eigenvector; and

and passing the classification feature vector through a classifier to obtain a classification result, wherein the classification result is used for indicating that the illumination power of the LED lamp belt at the current time point should be increased or decreased.

In the above method for controlling an LED strip, after respectively arranging the indoor luminance values and the outdoor luminance values of the plurality of predetermined time points including the current time point as input vectors, obtaining an indoor luminance feature vector and an outdoor luminance feature vector by a first sequence encoder including a one-dimensional convolution layer, the method includes: arranging the indoor brightness values of the plurality of preset time points including the current time point into an indoor brightness input vector according to a time dimension; performing full-concatenation encoding on the indoor luminance input vector using a full-concatenation layer of the first sequence encoder to extract high-dimensional implicit features of feature values of respective positions in the indoor luminance input vector according to the following formula:

represents a matrix multiplication; performing one-dimensional convolution encoding on the indoor luminance input vector by using a one-dimensional convolution layer of the first sequence encoder according to the following formula to extract each bit in the indoor luminance input vectorHigh-dimensional implicit associative features between the set feature values, wherein the formula is:

wherein, a is the width of the convolution kernel in the X direction, F is the parameter vector of the convolution kernel, G is the local vector matrix operated with the convolution kernel function, w is the size of the convolution kernel, and X represents the input vector; arranging the outdoor brightness values of the plurality of preset time points including the current time point into an outdoor brightness input vector according to a time dimension; performing full-concatenation encoding on the outdoor luminance input vector using a full-concatenation layer of the first sequence encoder to extract high-dimensional implicit features of feature values of respective positions in the outdoor luminance input vector according to the following formula:

In the above method for controlling an LED strip, calculating a luminance transfer matrix of the indoor luminance characteristic vector with respect to the outdoor luminance characteristic vector includes: calculating the brightness transfer matrix of the indoor brightness eigenvector relative to the outdoor brightness eigenvector by the following formula;

wherein the formula is:

M＝F ₁ *F ₂

In the method for controlling the LED strip, the step of obtaining the lighting power time sequence feature vector by passing the lighting power values of the LED strip at a plurality of predetermined time points including the current time point through a second sequence encoder including a one-dimensional convolutional layer includes: arranging the illumination power values of the LED lamp belts at a plurality of preset time points including the current time point into an illumination power input vector according to a time dimension; using a full-concatenation layer of the second sequence encoder to perform full-concatenation encoding on the illumination power input vector to extract high-dimensional implicit features of feature values of respective positions in the illumination power input vector, according to the following formula:

where X is the input vector, Y is the output vector, W is the weight matrix, B is the offset vector,

In the above method for controlling an LED strip, the correcting a feature value at each position in the illumination power timing feature vector to obtain a corrected illumination power timing feature vector includes: correcting the characteristic value of each position in the illumination power time sequence characteristic vector according to the following formula to obtain the corrected illumination power time sequence characteristic vector;

wherein the formula is:

In the method for controlling the LED light band, the correcting the characteristic values of the positions in the brightness transfer matrix to obtain a corrected brightness transfer matrix includes: correcting the characteristic value of each position in the brightness transfer matrix according to the following formula to obtain the corrected brightness transfer matrix;

wherein the formula is:

In the above LED strip control method, the d (v) is _i ，m _j,k ) Watch (A)And indicating Euclidean distances between the eigenvalues of each position in the illumination power time sequence eigenvector and the eigenvalue of each position in the brightness transfer matrix.

In the above method for controlling an LED strip, passing the classification feature vector through a classifier to obtain a classification result includes: processing the classification feature vector using the classifier to obtain the classification result with a formula: softmax { (W) _n ，B _n )：...：(W ₁ ，B ₁ ) I X }, wherein W ₁ To W _n As a weight matrix, B ₁ To B _n X is the classified feature vector.

Compared with the prior art, the LED lamp strip and the control method thereof provided by the application have the advantages that the indoor and outdoor brightness difference dynamic characteristics and the illumination power dynamic correlation characteristics of the LED lamp strip in the time sequence dimension are respectively excavated through the deep learning deep neural network model, whether the illumination power of the LED lamp strip at the current time point is properly classified is judged by utilizing the fusion characteristic information of the indoor and outdoor brightness difference dynamic characteristics and the illumination power dynamic correlation characteristics of the LED lamp strip, and in the characteristic fusion process, the overall distribution of the illumination power time sequence characteristic vector and the brightness transfer matrix approaches to an isotropic expression space with distinction degree relatively to each other through the comparison search space syntropy of characteristic values, so that the classification effect of the classification characteristic vector is enhanced. Therefore, the LED lamp strip can be adaptively adjusted according to the difference between the indoor brightness and the outdoor brightness, so that the requirements of people are met.

Drawings

The above and other objects, features and advantages of the present application will become more apparent by describing in more detail embodiments of the present application with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of the embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings, like reference numbers generally represent like parts or steps.

Fig. 1 is an application scene diagram of an LED strip according to an embodiment of the present application.

Fig. 2 is a block diagram of a LED strip according to an embodiment of the present application.

Fig. 3 is a block diagram of an ambient brightness data encoding module in an LED lamp strip according to an embodiment of the present application.

Fig. 4 is a flowchart of a method for controlling an LED strip according to an embodiment of the present application.

Fig. 5 is a schematic architecture diagram of a method for controlling an LED strip according to an embodiment of the present application.

Detailed Description

Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be understood that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and that the present application is not limited by the example embodiments described herein.

Overview of a scene

As mentioned above, at present, with the improvement of living standards of people, the requirement for beautifying living environment is higher and higher, and in order to beautify the living environment of people, the LED lamp strip with various light source colors is arranged in the hidden groove at the top of the room to form a beautiful light ring, which not only beautifies the living environment, but also brings a beautiful light to people at night.

However, the existing switch control systems of LED strips mostly use manually controlled electronic switches, which cannot automatically adjust the brightness of the LED strips according to the change of indoor brightness. Meanwhile, in the brightness adjustment process of the LED lamp strip, the difference between indoor brightness and outdoor brightness needs to be considered, so that the adjusted brightness of the LED lamp strip can meet the requirement of illumination, and the discomfort of human eyes caused by excessive stimulation on the human eyes can be avoided. Therefore, it is desirable to have an LED light strip that can be adaptively adjusted according to the difference between indoor brightness and outdoor brightness to meet the needs of people.

Accordingly, the inventor of the present application considers that when the LED strip is subjected to adaptive adjustment control, the difference between the indoor brightness and the outdoor brightness needs to be paid more attention to, so as to avoid the occurrence of a situation that human eyes feel uncomfortable due to an excessive difference between the indoor brightness and the outdoor brightness when the brightness of the LED strip is adjusted. This is also a categorical problem in nature, namely, whether to increase or decrease the lighting power of the LED strip at the current time point is adjusted by the brightness difference characteristics indoors and outdoors and the lighting power correlation characteristics of the LED strip.

Specifically, in the technical solution of the present application, first, an illumination power value, an indoor brightness value, and an outdoor brightness value of the LED strip at a plurality of predetermined time points including a current time point are obtained through each sensor deployed indoors and outdoors. Then, it should be understood that, considering that there is a relationship between the indoor luminance values and the outdoor luminance values at the plurality of predetermined time points including the current time point in the time dimension, respectively, in order to fully extract the hidden dynamically-changing relationship information, further arranging the indoor luminance values and the outdoor luminance values at the plurality of predetermined time points including the current time point as input vectors, respectively, to integrate luminance information for subsequent encoding processing, and then performing encoding processing on the input vectors arranged corresponding to the indoor luminance values and the outdoor luminance values, respectively, through a first sequence encoder including a one-dimensional convolution layer, to extract dynamically-changing implicit feature information of the indoor luminance values and the outdoor luminance values in time sequence, so as to obtain an indoor luminance feature vector and an outdoor luminance feature vector.

It should be understood that, since the indoor luminance variation characteristic and the outdoor luminance variation characteristic have a corresponding correlation relationship, in the technical solution of the present application, a luminance transfer matrix of the indoor luminance characteristic vector with respect to the outdoor luminance characteristic vector is further calculated to fuse the indoor luminance characteristic information and the outdoor luminance characteristic information.

Further, regarding the lighting power value of the LED strip, considering that the lighting power value also has a hidden association rule characteristic in a time sequence dimension, in the technical scheme of the present application, the lighting power value of the LED strip at a plurality of predetermined time points including the current time point is further encoded by a second sequence encoder including a one-dimensional convolution layer, so as to extract high-dimensional implicit characteristic information of the lighting power value of the LED strip at the plurality of predetermined time points and dynamic change association characteristic information of the lighting power value of the LED strip at the plurality of predetermined time points in the time sequence dimension, thereby obtaining a lighting power time sequence characteristic vector.

In this way, the illumination power time sequence feature vector and the brightness transfer matrix are subjected to matrix multiplication to fuse feature information of the illumination power time sequence feature vector and the brightness transfer matrix for classification, and then a classification result indicating that the illumination power of the LED lamp belt at the current time point should be increased or decreased is obtained.

However, when the illumination power time-series feature vector and the luminance transition matrix are subjected to matrix multiplication, it is difficult to avoid that distributed anisotropy exists at some positions when matrix multiplication is performed due to a difference in feature distribution between the time-series transition features of the illumination power as the time-series transition features of the luminance, and therefore, the illumination power time-series feature vector and the luminance transition matrix are corrected separately:

wherein v is _i And m _j，k Eigenvalues of the illumination power temporal eigenvector and the luminance transfer matrix, respectively, that are mapped into a probability space, d (v) _i ，m _j,k ) Representing the distance between the characteristic values and p being a hyperparameter, may initially be set as the mean of the distances between all characteristic values.

Here, the distributed anisotropy of some feature values may cause the feature distribution after the illumination power time-series feature vector is multiplied by the brightness transfer matrix to reside in a subset of an excessive discrete distribution in a high-dimensional feature space as a whole, so that the solution space of the classification feature vector with respect to the classification problem is degraded and the continuity of the feature distribution is lacked, and therefore, the search space homologization through the comparison of the feature values with each other makes the overall distribution of the illumination power time-series feature vector and the brightness transfer matrix approach to an isotropic and differentiated representation space with respect to each other, which enhances the classification effect of the classification feature vector and further improves the classification accuracy.

Based on this, the present application proposes a LED strip comprising: the illumination data and environment brightness data acquisition module is used for acquiring illumination power values, indoor brightness values and outdoor brightness values of the LED lamp belts at a plurality of preset time points including the current time point; the environment brightness data coding module is used for respectively arranging the indoor brightness values and the outdoor brightness values of the plurality of preset time points including the current time point into input vectors and then obtaining indoor brightness characteristic vectors and outdoor brightness characteristic vectors through a first sequence coder including one-dimensional convolution layers; the indoor and outdoor brightness data contrast coding module is used for calculating a brightness transfer matrix of the indoor brightness characteristic vector relative to the outdoor brightness characteristic vector; the illumination data coding module is used for enabling the illumination power values of the LED lamp strip at a plurality of preset time points including the current time point to pass through a second sequence encoder comprising a one-dimensional convolution layer so as to obtain an illumination power time sequence characteristic vector; the illumination characteristic correction module is used for correcting the characteristic value of each position in the illumination power time sequence characteristic vector to obtain a corrected illumination power time sequence characteristic vector; the brightness characteristic correction module is used for correcting the characteristic value of each position in the brightness transfer matrix to obtain a corrected brightness transfer matrix; the feature mapping fusion module is used for multiplying the corrected illumination power time sequence feature vector and the corrected brightness transfer matrix and mapping the high-dimensional brightness contrast change information of the corrected brightness transfer matrix to the high-dimensional feature space of the corrected illumination power time sequence feature vector to obtain a classification feature vector; and the illumination control result generation module is used for enabling the classified characteristic vectors to pass through a classifier to obtain a classification result, and the classification result is used for indicating that the illumination power of the LED lamp belt at the current time point should be increased or decreased.

Fig. 1 illustrates an application scenario diagram of an LED strip according to an embodiment of the present application. As shown in fig. 1, in this application scenario, first, the illumination power values, the indoor luminance values, and the outdoor luminance values of the LED strips (e.g., T as illustrated in fig. 1) at a plurality of predetermined time points including the current time point are acquired by respective sensors (e.g., a power sensor T1, an indoor luminance sensor T2, and an outdoor luminance sensor T3 as illustrated in fig. 1) disposed indoors and outdoors. Then, the obtained lighting power value, the indoor brightness value and the outdoor brightness value are input into a server (e.g., a cloud server S as illustrated in fig. 1) deployed with a LED strip algorithm, where the server can process the lighting power value, the indoor brightness value and the outdoor brightness value with the LED strip algorithm to generate a classification result indicating that the lighting power of the LED strip at the current time point should be increased or decreased.

Having described the general principles of the present application, various non-limiting embodiments of the present application will now be described with reference to the accompanying drawings.

Exemplary System

Fig. 2 illustrates a block diagram of a LED strip according to an embodiment of the present application. As shown in fig. 2, the LED strip 200 according to the embodiment of the present application includes: the illumination data and environment brightness data acquisition module 210 is configured to acquire illumination power values, indoor brightness values, and outdoor brightness values of the LED strips at a plurality of predetermined time points including a current time point; the environment brightness data encoding module 220 is configured to arrange the indoor brightness values and the outdoor brightness values of the multiple predetermined time points including the current time point into input vectors respectively, and then obtain an indoor brightness feature vector and an outdoor brightness feature vector through a first sequence encoder including a one-dimensional convolution layer; an indoor and outdoor brightness data contrast encoding module 230, configured to calculate a brightness transfer matrix of the indoor brightness eigenvector relative to the outdoor brightness eigenvector; the lighting data encoding module 240 is configured to pass the lighting power values of the LED strip at multiple predetermined time points including the current time point through a second sequence encoder including a one-dimensional convolutional layer to obtain a lighting power timing characteristic vector; an illumination characteristic correction module 250, configured to correct the characteristic value at each position in the illumination power time sequence characteristic vector to obtain a corrected illumination power time sequence characteristic vector; the brightness characteristic correction module 260 is configured to correct the characteristic values of the positions in the brightness transfer matrix to obtain a corrected brightness transfer matrix; a feature mapping fusion module 270, configured to multiply the corrected lighting power timing sequence feature vector with the corrected brightness transfer matrix, and map high-dimensional brightness contrast variation information of the corrected brightness transfer matrix into a high-dimensional feature space of the corrected lighting power timing sequence feature vector to obtain a classification feature vector; and an illumination control result generation module 280, configured to pass the classification feature vector through a classifier to obtain a classification result, where the classification result is used to indicate that the illumination power of the LED strip at the current time point should be increased or decreased.

Specifically, in this embodiment of the application, the illumination data and ambient brightness data collecting module 210 and the ambient brightness data encoding module 220 are configured to obtain an illumination power value, an indoor brightness value, and an outdoor brightness value of the LED strip at a plurality of predetermined time points including a current time point, arrange the indoor brightness value and the outdoor brightness value at the plurality of predetermined time points including the current time point into an input vector, and obtain an indoor brightness feature vector and an outdoor brightness feature vector through a first sequence encoder including a one-dimensional convolution layer. It should be understood that, when the adaptive adjustment and control is performed on the LED strip, attention needs to be paid to the difference between the indoor brightness and the outdoor brightness to avoid the occurrence of a situation that human eyes feel uncomfortable due to an excessive difference between the indoor brightness and the outdoor brightness when the brightness of the LED strip is adjusted. This is also a categorical problem in nature, namely, whether to increase or decrease the lighting power of the LED strip at the current time point is adjusted by the brightness difference characteristics indoors and outdoors and the lighting power correlation characteristics of the LED strip.

That is, specifically, in the technical solution of the present application, first, the illumination power value, the indoor luminance value, and the outdoor luminance value of the LED strip at a plurality of predetermined time points including the current time point are obtained by the respective sensors disposed indoors and outdoors. Then, it should be understood that, considering that there is a relationship between the indoor luminance values and the outdoor luminance values at the plurality of predetermined time points including the current time point in the time dimension, respectively, in the technical solution of the present application, in order to sufficiently extract such hidden dynamically-changing relationship information, the indoor luminance values and the outdoor luminance values at the plurality of predetermined time points including the current time point are further arranged as input vectors, respectively, so as to integrate luminance information for facilitating subsequent encoding processing. Then, the input vectors respectively corresponding to the indoor brightness values and the outdoor brightness values obtained through arrangement are subjected to coding processing in a first sequence coder comprising a one-dimensional convolution layer, so that the dynamic change implicit feature information of the indoor brightness values and the outdoor brightness values in time sequence is respectively extracted, and the indoor brightness feature vector and the outdoor brightness feature vector are obtained.

More specifically, in this embodiment of the application, the ambient brightness data encoding module includes: firstly, arranging the indoor brightness values of a plurality of preset time points including the current time point into an indoor brightness input vector according to a time dimension; performing full-concatenation encoding on the indoor luminance input vector using a full-concatenation layer of the first sequence encoder to extract high-dimensional implicit features of feature values of respective positions in the indoor luminance input vector according to the following formula:

represents a matrix multiplication; using the one-dimensional convolutional layer of the first sequence encoder as followsPerforming one-dimensional convolution coding on the indoor brightness input vector by using a formula to extract high-dimensional implicit correlation characteristics among characteristic values of all positions in the indoor brightness input vector, wherein the formula is as follows:

wherein, a is the width of the convolution kernel in the X direction, F is the parameter vector of the convolution kernel, G is the local vector matrix operated with the convolution kernel function, w is the size of the convolution kernel, and X represents the input vector. Then, arranging the outdoor brightness values of the plurality of preset time points including the current time point into an outdoor brightness input vector according to a time dimension; performing full-concatenation encoding on the outdoor luminance input vector using a full-concatenation layer of the first sequence encoder to extract high-dimensional implicit features of feature values of respective positions in the outdoor luminance input vector according to the following formula:

Fig. 3 illustrates a block diagram of an ambient brightness data encoding module in an LED strip according to an embodiment of the present application. As shown in fig. 3, the ambient brightness data encoding module 220 includes: an indoor luminance encoding unit 221, configured to arrange the indoor luminance values of the plurality of predetermined time points including the current time point into an indoor luminance input vector according to a time dimension; performing full-concatenation encoding on the indoor luminance input vector using a full-concatenation layer of the first sequence encoder to extract high-dimensional implicit features of feature values of respective positions in the indoor luminance input vector according to the following formula:

wherein, a is the width of the convolution kernel in the X direction, F is the parameter vector of the convolution kernel, G is the local vector matrix operated with the convolution kernel function, w is the size of the convolution kernel, and X represents the input vector; an outdoor brightness encoding unit 222, configured to arrange outdoor brightness values of the plurality of predetermined time points including the current time point into an outdoor brightness input vector according to a time dimension; using said first sequence encoderThe full-connection layer performs full-connection coding on the outdoor brightness input vector by using the following formula to extract high-dimensional implicit characteristics of characteristic values of all positions in the outdoor brightness input vector, wherein the formula is as follows:

represents a matrix multiplication; performing one-dimensional convolutional coding on the outdoor brightness input vector by using a one-dimensional convolutional layer of the first sequence encoder according to the following formula so as to extract high-dimensional implicit correlation features among feature values of all positions in the outdoor brightness input vector, wherein the formula is as follows:

Specifically, in this embodiment of the present application, the indoor and outdoor luminance data contrast encoding module 230 is configured to calculate a luminance transfer matrix of the indoor luminance eigenvector relative to the outdoor luminance eigenvector. It should be understood that, since the indoor brightness variation characteristic and the outdoor brightness variation characteristic have a corresponding correlation relationship, for example, outdoor brightness variation may affect indoor brightness measurement, in the technical solution of the present application, a brightness transfer matrix of the indoor brightness characteristic vector with respect to the outdoor brightness characteristic vector is further calculated to fuse the indoor brightness characteristic information and the outdoor brightness characteristic information.

More specifically, in this embodiment of the present application, the indoor and outdoor luminance data contrast encoding module is further configured to: calculating the brightness transfer matrix of the indoor brightness eigenvector relative to the outdoor brightness eigenvector by the following formula;

wherein the formula is:

M＝F ₁ *F ₂

Specifically, in this embodiment of the present application, the illumination data encoding module 240 is configured to pass the illumination power values of the LED strip at multiple predetermined time points including the current time point through a second sequence encoder including a one-dimensional convolutional layer to obtain an illumination power timing feature vector. It should be understood that, regarding the lighting power value of the LED strip, considering that the lighting power value also has a hidden association rule characteristic in a time sequence dimension, in the technical solution of the present application, the lighting power value of the LED strip at a plurality of predetermined time points including the current time point is further encoded by a second sequence encoder including a one-dimensional convolutional layer, so as to extract high-dimensional implicit characteristic information of the lighting power value of the LED strip at the plurality of predetermined time points and dynamic change association characteristic information of the lighting power value of the LED strip at the plurality of predetermined time points in the time sequence dimension, thereby obtaining a lighting power time sequence characteristic vector.

More specifically, in an embodiment of the present application, the lighting data encoding module is further configured to: arranging the illumination power values of the LED lamp belt at a plurality of preset time points including the current time point into an illumination power input vector according to a time dimension; using a full-concatenation layer of the second sequence encoder to perform full-concatenation encoding on the illumination power input vector to extract high-dimensional implicit features of feature values of respective positions in the illumination power input vector, according to the following formula:

Specifically, in the embodiment of the present application, the illumination characteristic correction module 250 and the brightness characteristic correction module 260 are configured to correct the characteristic value of each position in the illumination power time sequence characteristic vector to obtain a corrected illumination power time sequence characteristic vector, and correct the characteristic value of each position in the brightness transition matrix to obtain a corrected brightness transition matrix. It should be understood that after the illumination power timing characteristic vector and the brightness transfer matrix are obtained, the characteristics of the illumination power timing characteristic vector and the brightness transfer matrix are fused and classified, so that the LED strip can be intelligently and adaptively controlled. However, when the characteristics of the illumination power time-series characteristic vector and the luminance transition matrix are fused by matrix multiplication, it is difficult to avoid that distributed anisotropy exists at some positions when matrix multiplication is performed due to a characteristic distribution difference between the time-series transition characteristics of the illumination power and the time-series transition characteristics of the luminance, and therefore, in the technical solution of the present application, the illumination power time-series characteristic vector and the luminance transition matrix need to be corrected separately.

It should be understood that, here, the distribution anisotropy of some feature values may cause the feature distribution after multiplying the illumination power time sequence feature vector by the brightness transfer matrix to integrally reside in a subset of an excessive discrete distribution in a high-dimensional feature space, so that the solution space of the classification feature vector relative to the classification problem is degraded and the continuity of the feature distribution is lacked, therefore, the overall distribution of the illumination power time sequence feature vector and the brightness transfer matrix approaches an isotropic and differentiated representation space relative to each other by comparing the feature values with each other and searching the space syntropy, which enhances the classification effect of the classification feature vector and further improves the accuracy of the classification.

More specifically, in an embodiment of the present application, the illumination characteristic correction module is further configured to: correcting the characteristic value of each position in the illumination power time sequence characteristic vector by the following formula to obtain the corrected illumination power time sequence characteristic vector;

wherein the formula is:

wherein v is _i Feature values, m, representing respective positions of the illumination power time-sequential feature vector mapped to a probability space _j，k Characteristic values, d (v), representing the respective positions of the luminance transfer matrix mapped into the probability space _i ，m _j，k ) Representing the distance between the characteristic values and p being a hyperparameter, may initially be set as the mean of the distances between all characteristic values. In one specific example, the d (v) _i ，m _j,k ) And representing Euclidean distances between the characteristic value of each position in the illumination power time sequence characteristic vector and the characteristic value of each position in the brightness transfer matrix.

More specifically, in this embodiment of the application, the brightness characteristic correction module is further configured to: correcting the characteristic value of each position in the brightness transfer matrix according to the following formula to obtain the corrected brightness transfer matrix;

wherein the formula is:

wherein v is _i Feature values, m, representing respective positions of the illumination power time-sequential feature vector mapped to a probability space _i，k Characteristic values, d (v), representing the respective positions of the luminance transfer matrix mapped into the probability space _i ，m _j,k ) Representing the distance between the characteristic values and p being a hyperparameter, may initially be set as the mean of the distances between all characteristic values. In one specific example, the d (v) _i ，m _j,k ) And representing Euclidean distances between the eigenvalues of each position in the illumination power time sequence eigenvector and the eigenvalue of each position in the brightness transfer matrix.

Specifically, in this embodiment of the application, the feature mapping fusion module 270 and the illumination control result generation module 280 are configured to multiply the corrected illumination power timing feature vector and the corrected brightness transfer matrix, map the high-dimensional brightness contrast variation information of the corrected brightness transfer matrix into the high-dimensional feature space of the corrected illumination power timing feature vector to obtain a classification feature vector, and pass the classification feature vector through a classifier to obtain a classification result, where the classification result is used to indicate that the illumination power of the LED strip at the current time point should be increased or decreased. That is, in the technical solution of the present application, in order to fuse the corrected lighting power timing characteristic vector and the characteristic information of the corrected luminance transfer matrix, the corrected lighting power timing characteristic vector and the corrected luminance transfer matrix are further multiplied, and the high-dimensional luminance contrast variation information of the corrected luminance transfer matrix is mapped to the high-dimensional characteristic space of the corrected lighting power timing characteristic vector to obtain a classification characteristic vector for classification, so that a classification result indicating that the lighting power of the LED strip at the current time point should be increased or decreased can be obtained.

More particularly, in the present applicationIn an embodiment, the lighting control result generating module includes: processing the classification feature vector using the classifier to obtain the classification result with the following formula: softmax { (W) _n ，B _n )：...：(W ₁ ，B ₁ ) | X }, wherein W ₁ To W _n As a weight matrix, B ₁ To B _n Is a bias vector and X is the classification feature vector.

In summary, the LED strip 200 according to the embodiment of the present application is clarified, which respectively excavates the indoor and outdoor luminance difference dynamic features and the illumination power dynamic correlation features of the LED strip in the time sequence dimension through a deep neural network model of deep learning, and uses the fusion feature information of the two to classify whether the illumination power of the LED strip at the current time point is appropriate, and in the process of feature fusion, the overall distribution of the illumination power time sequence feature vector and the luminance transfer matrix approaches to an isotropic and differentiated expression space with respect to each other through the contrast search space syntropy of feature values, which enhances the classification effect of the classification feature vector. Therefore, the LED lamp strip can be adaptively adjusted according to the difference between the indoor brightness and the outdoor brightness, so that the requirements of people are met.

As described above, the LED strip 200 according to the embodiment of the present application may be implemented in various terminal devices, such as a server of an LED strip algorithm. In one example, the LED strip 200 according to the embodiments of the present application may be integrated into a terminal device as a software module and/or a hardware module. For example, the LED strip 200 may be a software module in the operating system of the terminal device, or may be an application developed for the terminal device; of course, the LED strip 200 may also be one of many hardware modules of the terminal device.

Alternatively, in another example, the LED strip 200 and the terminal device may be separate devices, and the LED strip 200 may be connected to the terminal device through a wired and/or wireless network and transmit the interactive information according to an agreed data format.

Exemplary method

Fig. 4 illustrates a flow chart of a method of controlling a LED strip. As shown in fig. 4, a method for controlling an LED strip according to an embodiment of the present application includes the steps of: s110, obtaining the illumination power value, the indoor brightness value and the outdoor brightness value of the LED lamp belt at a plurality of preset time points including the current time point; s120, respectively arranging the indoor brightness values and the outdoor brightness values of the plurality of preset time points including the current time point into input vectors, and then obtaining an indoor brightness characteristic vector and an outdoor brightness characteristic vector through a first sequence encoder including a one-dimensional convolution layer; s130, calculating a brightness transfer matrix of the indoor brightness characteristic vector relative to the outdoor brightness characteristic vector; s140, the illumination power values of the LED lamp strip at the plurality of preset time points including the current time point pass through a second sequence encoder including a one-dimensional convolution layer to obtain an illumination power time sequence characteristic vector; s150, correcting the characteristic value of each position in the illumination power time sequence characteristic vector to obtain a corrected illumination power time sequence characteristic vector; s160, correcting the characteristic value of each position in the brightness transfer matrix to obtain a corrected brightness transfer matrix; s170, multiplying the corrected illumination power time sequence eigenvector and the corrected brightness transfer matrix, and mapping the high-dimensional brightness contrast change information of the corrected brightness transfer matrix to the high-dimensional eigenvector space of the corrected illumination power time sequence eigenvector to obtain a classification eigenvector; and S180, passing the classified feature vectors through a classifier to obtain a classification result, wherein the classification result is used for indicating that the illumination power of the LED lamp strip at the current time point should be increased or decreased.

Fig. 5 illustrates an architecture diagram of a control method of an LED strip according to an embodiment of the present application. As shown IN fig. 5, IN the network architecture of the LED strip control method, firstly, after arranging the obtained indoor luminance values (e.g., IN1 as illustrated IN fig. 5) and outdoor luminance values (e.g., IN2 as illustrated IN fig. 5) of a plurality of predetermined time points including the current time point as input vectors (e.g., V1 and V2 as illustrated IN fig. 5), respectively, an indoor luminance feature vector (e.g., VF1 as illustrated IN fig. 5) and an outdoor luminance feature vector (e.g., VF2 as illustrated IN fig. 5) are obtained by a first sequence encoder (e.g., E1 as illustrated IN fig. 5) including one-dimensional convolutional layers; then, a luminance transfer matrix (e.g., MF1 as illustrated in fig. 5) of the indoor luminance feature vector with respect to the outdoor luminance feature vector is calculated; then, passing the obtained illumination power (e.g., IN as illustrated IN fig. 5) values of the LED strip at a plurality of predetermined time points including the current time point through a second sequence encoder (e.g., E2 as illustrated IN fig. 5) including one-dimensional convolution layers to obtain an illumination power time sequence feature vector (e.g., VC1 as illustrated IN fig. 5); then, correcting the eigenvalue of each position in the illumination power time sequence eigenvector to obtain a corrected illumination power time sequence eigenvector (for example, VC2 as illustrated in fig. 5); then, the eigenvalues of the respective positions in the luminance transfer matrix are corrected to obtain a corrected luminance transfer matrix (for example, MF2 as illustrated in fig. 5); then, multiplying the corrected illumination power timing eigenvector by the corrected brightness transfer matrix to map high-dimensional brightness contrast variation information of the corrected brightness transfer matrix into a high-dimensional eigenvector space of the corrected illumination power timing eigenvector to obtain a classification eigenvector (for example, VF as illustrated in fig. 5); and, finally, passing the classification feature vector through a classifier (e.g., a circle S as illustrated in fig. 5) to obtain a classification result, which is used to indicate that the illumination power of the LED light strip at the current time point should be increased or decreased.

More specifically, in steps S110 and S120, the illumination power values, the indoor brightness values and the outdoor brightness values of the LED strips at a plurality of predetermined time points including the current time point are obtained, and the indoor brightness values and the outdoor brightness values at the plurality of predetermined time points including the current time point are respectively arranged as input vectors, and then are passed through a first sequence encoder including one-dimensional convolution layers to obtain an indoor brightness feature vector and an outdoor brightness feature vector. It should be understood that, when the adaptive adjustment and control is performed on the LED strip, attention needs to be paid to the difference between the indoor brightness and the outdoor brightness to avoid the occurrence of a situation that human eyes feel uncomfortable due to an excessive difference between the indoor brightness and the outdoor brightness when the brightness of the LED strip is adjusted. This is also a categorical problem in nature, namely, whether to increase or decrease the lighting power of the LED strip at the current time point is adjusted by the brightness difference characteristics indoors and outdoors and the lighting power correlation characteristics of the LED strip.

That is, specifically, in the technical solution of the present application, first, the illumination power value, the indoor brightness value, and the outdoor brightness value of the LED strip at a plurality of predetermined time points including the current time point are obtained through respective sensors disposed indoors and outdoors. Then, it should be understood that, considering that there is a relationship between the indoor luminance values and the outdoor luminance values at the plurality of predetermined time points including the current time point in the time dimension, respectively, in the technical solution of the present application, in order to sufficiently extract such hidden dynamically-changing relationship information, the indoor luminance values and the outdoor luminance values at the plurality of predetermined time points including the current time point are further arranged as input vectors, respectively, so as to integrate luminance information for facilitating subsequent encoding processing. Then, the input vectors respectively corresponding to the indoor brightness values and the outdoor brightness values obtained through arrangement are subjected to coding processing through a first sequence coder comprising one-dimensional convolution layers, dynamic change implicit feature information of the indoor brightness values and the outdoor brightness values in time sequence is respectively extracted, and accordingly indoor brightness feature vectors and outdoor brightness feature vectors are obtained.

More specifically, in step S130, a luminance transfer matrix of the indoor luminance feature vector with respect to the outdoor luminance feature vector is calculated. It should be understood that, since the indoor brightness variation characteristic and the outdoor brightness variation characteristic have a corresponding correlation relationship, for example, outdoor brightness variation may affect indoor brightness measurement, in the technical solution of the present application, a brightness transfer matrix of the indoor brightness characteristic vector with respect to the outdoor brightness characteristic vector is further calculated to fuse the indoor brightness characteristic information and the outdoor brightness characteristic information.

More specifically, in step S140, the illumination power values of the LED strip at a plurality of predetermined time points including the current time point are passed through a second sequence encoder including a one-dimensional convolution layer to obtain an illumination power time sequence feature vector. It should be understood that, for the lighting power value of the LED strip, considering that the lighting power value also has a hidden association rule feature in a time sequence dimension, in the technical solution of the present application, the lighting power value of the LED strip at a plurality of predetermined time points including the current time point is further encoded by a second sequence encoder including a one-dimensional convolution layer, so as to extract high-dimensional implicit feature information of the lighting power value of the LED strip at the plurality of predetermined time points and dynamic change association feature information of the lighting power value of the LED strip at the plurality of predetermined time points in the time sequence dimension, thereby obtaining a lighting power time sequence feature vector.

More specifically, in steps S150 and S160, the eigenvalue of each position in the illumination power timing eigenvector is corrected to obtain a corrected illumination power timing eigenvector, and the eigenvalue of each position in the luminance transfer matrix is corrected to obtain a corrected luminance transfer matrix. It should be understood that, after the illumination power timing characteristic vector and the brightness transfer matrix are obtained, the characteristics of the two are fused and classified, so that the LED strip can be intelligently and adaptively controlled. However, when the characteristics of the illumination power time-series characteristic vector and the luminance transition matrix are fused by matrix multiplication, it is difficult to avoid that distributed anisotropy exists at some positions when matrix multiplication is performed due to a characteristic distribution difference between the time-series transition characteristics of the illumination power and the time-series transition characteristics of the luminance, and therefore, in the technical solution of the present application, the illumination power time-series characteristic vector and the luminance transition matrix need to be corrected separately.

More specifically, in steps S170 and S180, the corrected lighting power timing characteristic vector is multiplied by the corrected brightness transition matrix, the high-dimensional brightness contrast variation information of the corrected brightness transition matrix is mapped into the high-dimensional characteristic space of the corrected lighting power timing characteristic vector to obtain a classification characteristic vector, and the classification characteristic vector is passed through a classifier to obtain a classification result, wherein the classification result is used for indicating that the lighting power of the LED strip at the current time point should be increased or decreased. That is, in the technical solution of the present application, in order to fuse the corrected lighting power timing characteristic vector and the characteristic information of the corrected brightness transition matrix, the corrected lighting power timing characteristic vector and the corrected brightness transition matrix are further multiplied by mapping the high-dimensional brightness contrast variation information of the corrected brightness transition matrix to the high-dimensional characteristic space of the corrected lighting power timing characteristic vector to obtain a classification characteristic vector for classification, so as to obtain a classification result indicating that the lighting power of the LED strip at the current time point should be increased or decreased.

In summary, the method for controlling the LED strip according to the embodiment of the present application is clarified, which includes mining indoor and outdoor luminance difference dynamic features and illumination power dynamic correlation features of the LED strip in a time sequence dimension through a deep neural network model of deep learning, classifying whether the illumination power of the LED strip at a current time point is appropriate by using fusion feature information of the two, and enabling overall distributions of an illumination power time sequence feature vector and a luminance transfer matrix to approach an isotropic and differentiated expression space with respect to each other through comparison search space syntropy of feature values during feature fusion, so as to enhance a classification effect of classification feature vectors. Therefore, the LED lamp strip can be adaptively adjusted according to the difference between the indoor brightness and the outdoor brightness, so that the requirements of people are met.

The basic principles of the present application have been described above with reference to specific embodiments, but it should be noted that advantages, effects, etc. mentioned in the present application are only examples and are not limiting, and the advantages, effects, etc. must not be considered to be possessed by various embodiments of the present application. Furthermore, the foregoing disclosure of specific details is provided for purposes of illustration and understanding only, and is not intended to limit the application to the details which are set forth in order to provide a thorough understanding of the present application.

The block diagrams of devices, apparatuses, systems referred to in this application are only given as illustrative examples and are not intended to require or imply that the connections, arrangements, configurations, etc. must be made in the manner shown in the block diagrams. These devices, apparatuses, devices, systems may be connected, arranged, configured in any manner, as will be appreciated by those skilled in the art. Words such as "including," "comprising," "having," and the like are open-ended words that mean "including, but not limited to," and are used interchangeably therewith. As used herein, the words "or" and "refer to, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".

It should also be noted that in the devices, apparatuses, and methods of the present application, the components or steps may be decomposed and/or recombined. These decompositions and/or recombinations should be considered as equivalents of the present application.

The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

The foregoing description has been presented for purposes of illustration and description. Furthermore, the description is not intended to limit embodiments of the application to the form disclosed herein. While a number of example aspects and embodiments have been discussed above, those of skill in the art will recognize certain variations, modifications, alterations, additions and sub-combinations thereof.

Claims

1. A LED strip, comprising:

the illumination data coding module is used for enabling the illumination power values of the LED lamp strip at the plurality of preset time points including the current time point to pass through a second sequence encoder including a one-dimensional convolution layer so as to obtain an illumination power time sequence characteristic vector;

2. The LED light strip according to claim 1, wherein the ambient brightness data encoding module comprises:

the indoor brightness coding unit is used for arranging the indoor brightness values of the plurality of preset time points including the current time point into an indoor brightness input vector according to the time dimension; performing full-concatenation encoding on the indoor luminance input vector using a full-concatenation layer of the first sequence encoder to extract high-dimensional implicit features of feature values of respective positions in the indoor luminance input vector according to the following formula:

wherein, a is the width of the convolution kernel in the X direction, F is the parameter vector of the convolution kernel, G is the local vector matrix operated with the convolution kernel function, w is the size of the convolution kernel, and X represents the input vector;

the outdoor brightness coding unit is used for arranging the outdoor brightness values of the plurality of preset time points including the current time point into an outdoor brightness input vector according to a time dimension; fully-concatenate encoding the outdoor luminance input vector using the fully-concatenated layer of the first sequence encoder to extract the outdoor luminance input in the following formulaHigh-dimensional implicit features of feature values of respective positions in the vector, wherein the formula is:

3. The LED light strip according to claim 2, wherein the indoor and outdoor brightness data contrast coding module is further configured to: calculating the brightness transfer matrix of the indoor brightness eigenvector relative to the outdoor brightness eigenvector by the following formula;

wherein the formula is:

M＝F ₁ *F ₂

4. The LED light strip according to claim 3, wherein the illumination data encoding module is further configured to:

arranging the illumination power values of the LED lamp belts at a plurality of preset time points including the current time point into an illumination power input vector according to a time dimension;

using a full-concatenation layer of the second sequence encoder to perform full-concatenation encoding on the illumination power input vector to extract high-dimensional implicit features of feature values of respective positions in the illumination power input vector, according to the following formula:

represents a matrix multiplication;

performing one-dimensional convolution encoding on the illumination power input vector by using a one-dimensional convolution layer of the second sequence encoder according to the following formula to extract high-dimensional implicit correlation features among feature values of all positions in the illumination power input vector, wherein the formula is as follows:

5. The LED strip according to claim 4, wherein the lighting characteristic correction module is further configured to: correcting the characteristic value of each position in the illumination power time sequence characteristic vector according to the following formula to obtain the corrected illumination power time sequence characteristic vector;

wherein the formula is:

wherein v is _i Feature values, m, representing respective positions of the illumination power time-sequential feature vector mapped to a probability space _j,k Characteristic values, d (v), representing the respective positions of the luminance transfer matrix mapped into the probability space _i ,m _j,k ) Represents the distance between the characteristic values, and ρ is a hyperparameter.

6. The LED strip according to claim 5, wherein the brightness characteristic correction module is further configured to: correcting the characteristic value of each position in the brightness transfer matrix according to the following formula to obtain the corrected brightness transfer matrix;

wherein the formula is:

wherein v is _i Feature values, m, representing respective positions of the illumination power time-sequential feature vector mapped to a probability space _j,k Characteristic values, d (v), representing the respective positions of the luminance transfer matrix mapped into a probability space _i ,m _j,k ) Represents the distance between the characteristic values, and ρ is a hyperparameter.

7. LED strip according to claim 6, characterized in that said d (v) is _i ,m _j,k ) And representing Euclidean distances between the eigenvalues of each position in the illumination power time sequence eigenvector and the eigenvalue of each position in the brightness transfer matrix.

8. The LED light strip according to claim 7, wherein the lighting control result generation module is further configured to: processing the classification feature vector using the classifier to obtain the classification result with a formula: softmax { (W) _n ,B _n )：…：(W ₁ ,B ₁ ) | X }, wherein W ₁ To W _n As a weight matrix, B ₁ To B _n Is a bias vector and X is the classification feature vector.

9. A control method of an LED lamp belt is characterized by comprising the following steps:

the lighting power values of the LED lamp strip at a plurality of preset time points including the current time point pass through a second sequence encoder including a one-dimensional convolution layer to obtain a lighting power time sequence characteristic vector;

and passing the classified feature vectors through a classifier to obtain a classification result, wherein the classification result is used for indicating that the illumination power of the LED lamp strip at the current time point should be increased or decreased.

10. The method for controlling the LED light strip according to claim 9, wherein calculating the luminance transfer matrix of the indoor luminance eigenvector with respect to the outdoor luminance eigenvector comprises:

calculating the brightness transfer matrix of the indoor brightness eigenvector relative to the outdoor brightness eigenvector by the following formula;

wherein the formula is:

M＝F ₁ *F ₂