Remote programmable IC light and color modulation RGB colorful LED light source

Abstract

The invention discloses a remotely programmable IC dimming and color adjusting RGB colorful LED light source, relates to the technical field of LED light source control, and solves the technical problem that the actual lighting effect of the LED light source is inconsistent with the expected effect due to the fact that the influence of external environmental factors on the overall lighting effect is neglected in the prior art. The behavior data of the target user and the environmental data of the LED light source are identified by constructing the parameter prediction model so as to obtain the light source control parameters suitable for the target user, and the LED light source is controlled through the light source control parameters so as to eliminate the influence of environmental factors on the illumination effect of the LED light source; the input of a neural network II in the parameter prediction model is environmental data and a behavior label, and the output of the neural network II is a light source control parameter of the LED light source required by a target user under the environmental data; the influence of environmental factors is considered during prediction of the light source control parameters, so that the light source characteristics of the LED light source meet the requirements of a target user.

Description

Technical Field

[0001] This application belongs to the technical field of LED light source control, and specifically relates to a remotely programmable IC dimming and color - tuning RGB multi - color LED light source. Background Art

[0002] An IC dimming and color - tuning RGB multi - color LED light source is an LED lighting product integrated with intelligent control functions. It realizes the brightness adjustment (dimming) and color change (color - tuning) of LED lights through a built - in IC chip, and can provide various color and brightness combinations to meet the lighting needs of different scenarios.

[0003] LED light sources are widely used in various scenarios, such as home lighting, commercial lighting, landscape lighting, etc., especially in home lighting scenarios, which are deeply integrated with smart home technology. In the field of smart home, first, user instructions are obtained, the user instructions are analyzed to determine the control rules for the LED light source, and then the brightness, color, etc. of the LED light source are automatically controlled according to the control rules. When the existing LED light source control process automatically controls the LED light source, it only focuses on the control of the LED light source itself, while ignoring the influence of external environmental factors on the overall lighting effect, resulting in the actual lighting effect of the LED light source being inconsistent with the lighting effect required by the control rules, thus affecting the user experience.

[0004] This application provides a remotely programmable IC dimming and color - tuning RGB multi - color LED light source to solve the above - mentioned technical problems. Summary of the Invention

[0005] This application aims to solve at least one of the technical problems existing in the prior art; for this purpose, this application proposes a remotely programmable IC dimming and color - tuning RGB multi - color LED light source, which is used to solve the technical problem that the prior art ignores the influence of external environmental factors on the overall lighting effect, resulting in the actual lighting effect of the LED light source being inconsistent with the expected effect.

[0006] To achieve the above object, the first aspect of this application provides a remotely programmable IC dimming and color - tuning RGB multi - color LED light source, including a central control module, a data acquisition module connected thereto, and at least one LED light source;

[0007] Data acquisition module: used to collect the behavior data of the target user and the environmental data of the location where the LED light source is located through the set data sensors; among them, the behavior data includes time, location, and actions, and the environmental data includes environmental temperature, environmental humidity, and environmental brightness;

[0008] Central control module: used to build a parameter prediction model based on an artificial intelligence model; after preprocessing the behavior data and environmental data, input them into the parameter prediction model to obtain light source control parameters; wherein, the artificial intelligence model is a cascaded structure including neural network one and neural network two; and,

[0009] used to control the LED light source through the light source control parameters; wherein, the light source control parameters are used to control the brightness, color temperature and color of the LED light source.

[0010] Preferably, building a parameter prediction model based on an artificial intelligence model includes:

[0011] Extract environmental historical data, training data one and corresponding historical light source parameters from the standard training data; wherein, training data one includes behavior data and behavior labels;

[0012] Build an artificial intelligence model; use the environmental historical data and training data one to train neural network one, use the environmental historical data, the output parameters of neural network one and the historical light source parameters to train neural network two, and mark the trained artificial intelligence model as the parameter prediction model.

[0013] Preferably, the standard training data is extracted from the historical record data, including:

[0014] Record the behavior of the target user through the data sensor to obtain historical record data; extract the behavior data and environmental historical data of the target user from the historical record data, set behavior labels according to the behavior data and the environmental data; integrate the behavior data and behavior labels into training data one;

[0015] Extract the light source control parameters corresponding to training data one from the historical record data, mark them as historical light source parameters; integrate the environmental historical data, training data one and historical light source parameters into standard training data.

[0016] Preferably, recording the behavior of the target user through the data sensor includes:

[0017] Collect the behavior data of the target user through the data sensor, conduct a preliminary classification of the behavior of the target user based on the behavior data to obtain a preliminary classification result;

[0018] Wait for the target user to confirm or adjust the preliminary classification result to generate a behavior label.

[0019] Preferably, after generating the behavior label, determining the historical light source parameters according to the behavior label includes:

[0020] Retrieve the standard light source features corresponding to the behavior label; wherein, the standard light source features include brightness, color temperature and color, and the standard light source features are preset and stored in the central control module;

[0021] Control the LED light source according to the standard light source characteristics, and adjust the control result of the LED light source according to the feedback result of the target user to obtain the target light source characteristics;

[0022] Extract the light source control parameters corresponding to the target light source characteristics and integrate them into historical light source parameters; among them, the light source control parameters include voltage, current or RBG ratio.

[0023] Preferably, adjusting the control result of the LED light source according to the feedback result of the target user includes:

[0024] After the control of the LED light source is completed, map the standard light source characteristics to the intelligent terminal of the target user; among them, the intelligent terminal includes a smart phone or a tablet computer;

[0025] The target user adjusts the standard light source characteristics through the intelligent terminal in combination with their own experience, and generates a feedback result from the adjustment content of the target user.

[0026] Preferably, adjust the LED light source according to the feedback result, and map the adjusted light source characteristics to the intelligent terminal of the target user;

[0027] After the target user completes the adjustment, record the light source characteristics corresponding to the LED light source as the target light source characteristics.

[0028] Preferably, extracting the light source control parameters corresponding to the target light source characteristics includes:

[0029] Taking the LED lamp bead at the center position of the LED light source as the reference lamp bead, numbering the LED lamp beads along a set track based on the reference lamp bead to obtain a lamp bead label; among them, the set track includes a counterclockwise track, a clockwise track or a linear track;

[0030] Extract the light source control parameters of several LED lamp beads under the target light source characteristics, associate the light source control parameters with the lamp bead labels of the corresponding LED lamp beads, and integrate and generate historical light source parameters in the order of the lamp bead labels.

[0031] Compared with the prior art, the beneficial effects of this application are:

[0032] 1. When the present application conducts intelligent control on the LED light source, by constructing a parameter prediction model to identify the behavior data of the target user and the environmental data of the LED light source to obtain the light source control parameters suitable for the target user, and controlling the LED light source through the light source control parameters to eliminate the influence of environmental factors on the lighting effect of the LED light source; the input of neural network two in the parameter prediction model is environmental data and behavior labels, and the behavior labels include the brightness, color temperature, and color required by the target user under this environmental data, and its output is the light source control parameters of the LED light source when the target user's requirements are to be met under this environmental data, that is, current, voltage, etc.; the influence of environmental factors has been considered during the prediction of the light source control parameters. Therefore, when controlling the LED light source according to the light source control parameters, the light source characteristics of the LED light source meet the requirements of the target user.

[0033] 2. During the testing process of the present application, the light source control parameters extracted can be the light source control parameters corresponding to each LED lamp bead in the LED light source. Then, based on the light source control parameters of several LED lamp beads, the zonal control or even single control of the LED light source can be realized, which can improve the accuracy of LED light source control; on this basis, the timing control of the light source control parameters of the LED lamp beads can also be carried out to create different atmospheres. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0035] Figure 1 Schematic diagram of the control method of the IC dimming and color mixing RGB multi-color LED light source in Embodiment 1 of the present application;

[0036] Figure 2 Schematic diagram of the control system of the IC dimming and color mixing RGB multi-color LED light source in Embodiment 1 of the present application;

[0037] Figure 3 Schematic diagram of the cascade structure of the artificial intelligence model in Embodiment 1 of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0038] The following will clearly and completely describe the technical solutions of the present application in combination with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0039] In the field of smart home, the intelligent adjustment of LED light sources is a very important content. When adjusting the LED light source, various factors that may affect the LED light source will be comprehensively considered, such as ambient light intensity or the scene where it is located. Therefore, in the prior art during the adjustment process of the LED light source, it is necessary to first obtain the control instruction sent by the user, determine the scene where the user is located through this control instruction, and then combine the environmental data where the user is located to determine the control parameters of the LED light source, thereby realizing the intelligent control of the LED light source.

[0040] The existing solutions need to first obtain the control instruction sent by the user. This control instruction is generally selected and generated by the user through a smart terminal, that is, the user needs to intervene first every time the LED light source is controlled, and intelligent control cannot be achieved. Moreover, when determining the control parameters of the LED light source based on the scene and environmental data, the scene and environmental data are input into a pre-trained artificial intelligence model after preprocessing to determine the corresponding control parameters. However, when training the artificial intelligence model, the corresponding standard control parameters are obtained in different environments in different scenes. This standard control parameter can only control the LED light source to achieve a preset ideal effect, but this ideal effect may not be suitable for the user, resulting in poor control effects.

[0041] Embodiment 1:

[0042] Please refer to Figure 1 - Figure 2 , the first aspect embodiment of this application provides a remotely programmable IC dimming and color mixing RGB multi-color LED light source, including a central control module, a data acquisition module connected thereto, and at least one LED light source;

[0043] Data acquisition module: used to collect the behavior data of the target user and the environmental data of the location where the LED light source is located through the set data sensors;

[0044] Central control module: used to construct a parameter prediction model based on the artificial intelligence model; after preprocessing the behavior data and environmental data, input them into the parameter prediction model to obtain the light source control parameters; and, used to control the LED light source through the light source control parameters.

[0045] The IC dimming and color mixing RGB multi-color LED light source is an LED lighting product integrated with intelligent control functions. It realizes the brightness adjustment (dimming) and color change (color mixing) of the LED light through the built-in IC chip, and can provide a variety of color and brightness combinations to meet the lighting needs of different scenes.

[0046] The central control module (equivalent to a controller) in this example can be used to remotely control the LED light source. The central control module is communicatively and / or electrically connected to the data acquisition module and the LED light source. The data acquisition module is responsible for collecting the data required during the control process of the LED light source through data sensors, including behavior data for identifying user behavior and environmental data affecting the control effect of the LED light source.

[0047] In this embodiment, the positions for collecting the behavior data and the environmental data are the same position. It can also be understood that the target user and the LED light source to be controlled are in the same area, such as in a study or a gym. Therefore, the environmental data is actually the environmental data of the same area. It should be noted that if the target user and the LED light source are in two places, such as the target user is in the living room and the LED light source is in the study, and the target user is walking towards the study with a book in hand, then the behavior data and the environmental data corresponding to the LED light source are different at this time.

[0048] The data sensors are mainly used to collect behavior data and environmental data, which can be realized by various types of sensors related to smart home, such as temperature sensors, humidity sensors, cameras, etc. In this embodiment, the setting of the data sensors mainly serves the intelligent control of the LED light source. Therefore, the setting location is determined according to the data to be collected, but user authorization or respect for user privacy is required during the setting process. The behavior data includes time, location, and action. This behavior data is mainly used to identify the corresponding scenario of the target user. At the same time, environmental data can also be introduced when judging the scenario, that is, in a certain environment, if the target user makes a certain action at a certain time and at a certain location, it can be determined that the user is in a certain scenario or wants to enter a certain scenario. The environmental data includes environmental temperature, environmental humidity, and environmental brightness. This environmental data is mainly used for the intelligent control of the LED light source, but also for assisting in identifying the corresponding scenario of the target user. Therefore, if the target user and the LED light source are in different scenarios, the environmental data needs to be collected separately.

[0049] The implementation logic of this embodiment can be summarized as follows: When the target user and the LED light source are in the same area, automatically identify the behavior data of the target user and the environmental data where the LED light source is located; use the environmental data and the behavior data to identify the behavior of the target user, and then integrate and process the behavior label corresponding to this behavior with the environmental data to identify the light source control parameters of the LED light source. It should be noted that if the target user and the LED light source are in two areas, the environmental data needs to be collected separately, and the above logic basically remains unchanged.

[0050] Therefore, the construction of the parameter prediction model in this example is very important. The steps for constructing the parameter prediction model based on an artificial intelligence model can be referred to as follows, including:

[0051] Extract environmental historical data, training data one, and corresponding historical light source parameters from standard training data; construct an artificial intelligence model; use the environmental historical data and training data one to train neural network one, use the environmental historical data, the output parameters of neural network one, and the historical light source parameters to train neural network two, and mark the trained artificial intelligence model as a parameter prediction model.

[0052] The standard training data is extracted from historical record data, and the historical record data is mainly the data generated by the behavior of the target user recorded by the data sensor. To ensure the reliability and processing efficiency of the historical record data, the behavior data of the target user is collected in real time by the data sensor during the recording period, mainly including the collection time, actions, and location, etc. Then, a classification model (such as a support vector machine model) is used to classify the behavior data to obtain a preliminary classification result. This preliminary classification result is only a simple classification. The target user confirms the preliminary classification result. If it is inconsistent with the actual intention of the target user, the preliminary classification result is adjusted. When all the preliminary classification results meet the requirements of the target user, behavior labels are generated according to the preliminary classification results.

[0053] Exemplarily, the acquisition of the preliminary classification result can be understood with reference to the following example:

[0054] 1) The behavior data of the target user is collected in real time by the data sensor during the recording period, mainly including the collection time, actions, and location, etc. For example, motion sensors, infrared sensors, and light sensors can be used to collect the user's behavior data and environmental data.

[0055] Collection time: Record the user's activities at different time periods, such as 8 am, 4 pm, 10 pm, etc.

[0056] Actions: Record the user's actions, such as reading, watching TV, sleeping, exercising, etc.

[0057] Location: Record the user's location in the room, such as the living room, bedroom, kitchen, etc.

[0058] 2) Normalize the collected data so that it is on the same order of magnitude for easy model calculation. For example, the minimum-maximum normalization method can be used to scale data such as temperature, humidity, and brightness to the [0,1] interval. For the user's activity time, it can be converted into a time point in a day. For example, convert 8 am to 0.333 and 4 pm to 0.667.

[0059] It should be noted that when obtaining the logic of standard training data in this embodiment, the behavior data of the target user is recorded in real time. After accumulating a certain amount of behavior data, the behavior data is classified by a classification model. The initial classification result can be understood as determining the actions corresponding to the behavior data, such as reading, resting, fitness, entertainment, etc. Then, the actions corresponding to each behavior data determined by the target user are determined and adjusted, and then behavior labels are set for each behavior data. At this time, each behavior data will correspond to a behavior label.

[0060] The above classification model is used to identify what specific actions the behavior data of the target user belongs to, so as to set subsequent behavior labels. This classification model can be implemented based on a support vector machine. Using a support vector machine for data classification has been disclosed in existing solutions and can be implemented with reference to existing solutions.

[0061] After determining the behavior labels corresponding to each behavior data of the target user, the light source control parameters are set and adjusted according to the behavior labels. Specifically, the preset standard light source features corresponding to the behavior labels are retrieved. The standard light source features include brightness, color temperature, and color; the LED light source is controlled according to the light source control parameters corresponding to the standard light source features, and then the target user adjusts according to their own experience of the LED light source, and the light source control parameters corresponding to the adjusted target light source features are used as historical light source parameters.

[0062] The target light source is obtained during the continuous adjustment and learning process. Specifically, after controlling the LED light source according to the light source control parameters corresponding to the standard light source features, the standard light source features are mapped to the intelligent terminal of the target user. If the target user is not satisfied with the actual performance of the LED light source, the standard light source features are finely adjusted through the intelligent terminal until the effect of the LED light source meets the requirements of the target user. At this time, the standard light source features that the target user is satisfied with are used as the target light source features. It may take multiple rounds of adjustment to satisfy the target user.

[0063] It should be clear here that the standard light source characteristics are the light source characteristics that meet the public's needs in several scenarios built in by the manufacturer, but they do not meet everyone's needs. Therefore, in the same scenario, if controlling the LED light source according to the standard light source characteristics cannot meet the requirements of the target user, the adjustment process of the target user is equivalent to a custom setting. Furthermore, if the target user's requirements are exactly the same as the public's requirements, that is, theoretically, controlling the LED light source according to the standard light source characteristics can meet the requirements of the target user, but it may be affected by the environment where the LED light source is located, such as ambient light intensity, ambient temperature, ambient humidity, etc., resulting in a deviation between the actual effect of the LED light source and the test result. At this time, the adjustment process of the target user to the LED light source is equivalent to eliminating the error caused by environmental changes. Whether it is a custom setting or eliminating the error caused by environmental changes, it is to improve the experience of the target user. Moreover, during the adjustment process, it is also necessary to pay attention to setting a reasonable range for the adjustment of the standard light source characteristics to avoid damaging the target user due to the actual effect of the LED light source after adjustment.

[0064] In this way, the behavior labels are first determined based on the behavior data of the target user, and then the target light source characteristics most suitable for the target user under each behavior label are determined. The light source control parameters corresponding to the target light source characteristics are the most preferred LED light source control effects of the target user in this environment and scenario. In this way, a piece of data can be obtained, including environmental historical data, behavior data, behavior labels, and historical light source parameters. By continuously collecting and adjusting, multiple pieces of data can be obtained, that is, standard training data.

[0065] It should be noted that since the target user and the lighting scenario are relatively simple, even if all tests are carried out, it will not take too much time and cost, and the data obtained from the target user's tests is more targeted. Compared with the built-in standard light source characteristics and their corresponding light source control parameters, the historical light source parameters adjusted according to the feedback of the target user are more suitable for the target user in the same scenario and environment.

[0066] After obtaining the standard training data, the artificial intelligence model can be trained or updated. The training process of the artificial intelligence model can refer to the following steps:

[0067] Extract environmental historical data, training data one, and the corresponding historical light source parameters from the standard training data; among them, training data one includes behavior data and behavior labels;

[0068] Construct an artificial intelligence model; use the environmental historical data and training data one to train neural network one, use the environmental historical data, the output parameters of neural network one, and the historical light source parameters to train neural network two, and mark the trained artificial intelligence model as a parameter prediction model.

[0069] When training an artificial intelligence model, environmental historical data and behavioral data are used as the inputs of Neural Network 1, and behavioral labels are used as the outputs for training. Then, the outputs of Neural Network 1, i.e., behavioral labels, and environmental historical data are used as the inputs of Neural Network 2, and historical light source parameters are used as the outputs for training, finally obtaining a trained artificial intelligence model. Neural Network 1 mainly conducts data classification and can be a BP neural network model or a support vector machine model. Neural Network 2 is mainly responsible for constructing mapping relationships and can be a BP neural network model or an RBF neural network model. Please refer to Figure 3 Neural Network 1 and Neural Network construct an artificial intelligence model through a cascaded structure.

[0070] Exemplarily, the specific structure of the artificial intelligence model can be referred to as follows:

[0071] 1. Neural Network 1

[0072] Input nodes of the input layer: Environmental historical data and behavioral data; Number of nodes: Assume that the environmental historical data has 3 features (temperature, humidity, brightness) and the behavioral data has 3 features (time, location, action), then the input layer has 6 nodes.

[0073] Hidden layer nodes of the hidden layer: Assume that the hidden layer has 12 nodes and the ReLU activation function is used.

[0074] Output nodes of the output layer: Behavioral labels; Number of nodes: Assume that the behavioral labels have 4 categories (reading, entertainment, rest, fitness), then the output layer has 4 nodes and the Softmax activation function is used.

[0075] 2. Neural Network 2

[0076] Input nodes of the input layer: The outputs of Neural Network 1 (behavioral labels) and environmental historical data; Number of nodes: Assume that the behavioral labels have 4 features (reading, entertainment, rest, fitness) and the environmental historical data has 3 features (temperature, humidity, brightness), then the input layer has 7 nodes.

[0077] Hidden layer nodes of the hidden layer: Assume that the hidden layer has 12 nodes and the ReLU activation function is used.

[0078] Output nodes of the output layer: Historical light source parameters; Number of nodes: Assume that the historical light source parameters have 3 (brightness, color temperature, color), then the output layer has 3 nodes and the linear activation function is used.

[0079] The trained artificial intelligence model is marked as a parameter prediction model. The behavior data and environmental data of the target user are collected in real time through a data sensor and input into the parameter prediction model. The neural network in the parameter prediction model preprocesses the environmental data and behavior data, such as normalization, and outputs the corresponding behavior labels. Then, the environmental data and behavior labels are normalized to output the light source control parameters suitable for the target user. The LED light source can be adjusted in a timely manner according to the light source control parameters.

[0080] Embodiment 2: Compared with Embodiment 1, the difference of this embodiment from the embodiment is that the artificial intelligence model can be trained using the test data of multiple test users.

[0081] The acquisition method of the standard training data is the same as that in Embodiment 1, that is, the test user is allowed to make actions that conform to the preset scenarios (reading, resting, entertaining, fitness, etc.) at will. The behavior data of the test user is collected through a data sensor, and then the initial classification result is obtained by identifying the behavior data through a classification model. The test user confirms or adjusts the initial classification result to generate behavior labels. Then, the LED light source is adjusted according to the behavior labels, and continuous adjustment is made according to the feedback results of the test user to obtain the corresponding light source control parameters. The environmental data, behavior data, and light source control parameters of the environment where the test user is located are integrated into one piece of data. Then, one test user can obtain multiple pieces of data, and several test users can obtain more test data.

[0082] When training the artificial intelligence model, the identity label of the test user is inserted after the data corresponding to it, and the trained artificial intelligence model is marked as a parameter prediction model. The identity labels mainly include gender, age, hobbies (reading, entertainment, fitness, etc.), myopia degree, astigmatism degree, hyperopia degree, etc. When determining the light source control parameters required by the target user, the identity label of the target user can be inserted into the corresponding environmental data and behavior data to obtain the light source control parameters that suit the user himself.

[0083] Since the artificial intelligence model trained with the data obtained by testing several different test users to obtain a parameter prediction model can improve the practicability and generalization ability of the artificial intelligence model. The parameter prediction model is built into the central control module, and then the central control module is arranged in the cloud server, which further improves the application range of the LED light source, and it is not necessary to test the target user, making the intelligent control simpler.

[0084] Embodiment 3: Compared with Embodiment 1, the light source control parameters in this embodiment are not a set of data, but multiple sets of data, and each set of data is used to control one LED lamp bead in the LED light source.

[0085] Taking the LED lamp bead at the center position of the LED light source as the reference lamp bead, numbering the LED lamp beads according to the set trajectory based on the reference lamp bead to obtain the lamp bead identifier; extracting the light source control parameters of several LED lamp beads under the target light source characteristics, associating the light source control parameters with the lamp bead identifiers of the corresponding LED lamp beads, and integrating and generating the historical light source parameters in the order of the lamp bead identifiers.

[0086] During the testing process for both test users and target users, after determining the light source characteristics that meet the users, it is necessary to extract the light source characteristic parameters corresponding to the light source characteristics. If the LED light source has only one LED lamp bead, only the official control parameters of this LED lamp bead need to be extracted. If the LED light source includes multiple LED lamp beads, the LED lamp beads are marked to obtain the light source control parameters of each LED lamp bead. At this time, the light source control parameters are composed of multiple arrays. Constructing the historical light source parameters with the light source characteristic parameters of a single LED lamp bead enables the trained parameter prediction model to also output multiple sets of parameters. According to the multiple sets of output parameters, the zoned control of the LED light source can be realized, and even individual control can be achieved, enabling the refined control of the LED light source.

[0087] The above set trajectory includes a counterclockwise trajectory, a clockwise trajectory, or a linear trajectory. For a circular LED light source, a counterclockwise trajectory or a clockwise trajectory can be adopted. For a rectangular LED light source, a linear trajectory is adopted. The lamp bead identifier can be a digital number, used to distinguish the arrays corresponding to each LED lamp bead.

[0088] The above embodiments are only used to illustrate the technical method of this application and not to limit it. Although this application has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of this application can be modified or equivalently replaced without departing from the spirit and scope of the technical method of this application.

Claims

1. A remotely programmable IC dimming and color adjustment RGB multi-color LED light source, characterized in that: It includes a central control module, a data acquisition module connected thereto, and at least one LED light source; Data collection module: used to collect the behavior data of the target user and the environmental data of the location of the LED light source through the set data sensor; wherein the behavior data includes time, location and action, and the environmental data includes environmental temperature, environmental humidity and environmental brightness; Central control module: used to build a parameter prediction model based on an artificial intelligence model; after preprocessing the behavior data and the environmental data, input them into the parameter prediction model to obtain light source control parameters; wherein the artificial intelligence model is a cascade structure including a neural network 1 and a neural network 2; and, Used to control the LED light source through the light source control parameters; wherein the light source control parameters are used to control the brightness, color temperature and color of the LED light source.

2. The remotely programmable IC dimming and color-adjusting RGB multi-color LED light source according to claim 1, characterized in that: The parameter prediction model is constructed based on the artificial intelligence model, including: Extracting environmental historical data, training data one and corresponding historical light source parameters from standard training data; wherein the training data one includes behavior data and behavior labels; Construct an artificial intelligence model; use the environmental historical data and the training data to train neural network one, use the environmental historical data, the output parameters of neural network one and the historical light source parameters to train neural network two, and mark the trained artificial intelligence model as a parameter prediction model.

3. The remotely programmable IC dimming and color-adjusting RGB multi-color LED light source according to claim 2, characterized in that: The standard training data is extracted from historical record data and includes: Record the behavior of the target user by the data sensor to obtain historical record data; extract the behavior data of the target user and the environmental history data from the historical record data, and set a behavior label according to the behavior data and the environmental data; integrate the behavior data and the behavior label into training data 1; The light source control parameters corresponding to the training data 1 are extracted from the historical record data and marked as historical light source parameters; the environmental historical data, training data 1 and historical light source parameters are integrated into standard training data.

4. The remotely programmable IC dimming and color-adjusting RGB multi-color LED light source according to claim 2, characterized in that: The target user's behavior is recorded by the data sensor, including: Collecting the behavior data of the target user through the data sensor, and performing preliminary scene classification on the behavior of the target user based on the behavior data to obtain a preliminary classification result; After the target user confirms or adjusts the initial classification result, a behavior label is generated.

5. The remotely programmable IC dimming and color-adjusting RGB multi-color LED light source according to claim 4, characterized in that: After the behavior tag is generated, determining the historical light source parameter according to the behavior tag includes: Retrieving the standard light source characteristics corresponding to the behavior tag; wherein the standard light source characteristics include brightness, color temperature and color, and the standard light source characteristics are pre-set and stored in the central control module; The LED light source is controlled according to the standard light source characteristics, and the control result of the LED light source is adjusted according to the feedback result of the target user to obtain the target light source characteristics; The light source control parameters corresponding to the target light source characteristics are extracted and integrated into historical light source parameters; wherein the light source control parameters include voltage, current or RBG ratio.

6. The remotely programmable IC dimming and color-adjusting RGB multi-color LED light source according to claim 5, characterized in that: The control result of the LED light source is adjusted according to the feedback result of the target user, including: After the LED light source control is completed, the standard light source characteristics are mapped to the target user's smart terminal; wherein the smart terminal includes a smart phone or a tablet computer; The target user adjusts the standard light source characteristics through the smart terminal based on his own experience, and the adjustment content of the target user is generated into a feedback result.

7. The remotely programmable IC dimming and color-adjusting RGB multi-color LED light source according to claim 6, characterized in that: Adjusting the LED light source according to the feedback result, and mapping the adjusted light source characteristics to the smart terminal of the target user; After the target user completes the adjustment, the light source characteristics corresponding to the LED light source are recorded as target light source characteristics.

8. The remotely programmable IC dimming and color-adjusting RGB multi-color LED light source according to claim 5, characterized in that: Extracting light source control parameters corresponding to the target light source characteristics includes: The LED lamp bead at the center of the LED light source is used as a reference lamp bead, and the LED lamp beads are numbered according to a set trajectory based on the reference lamp bead to obtain a lamp bead mark; The light source control parameters of the plurality of LED lamp beads under the target light source characteristics are extracted, the light source control parameters are associated with the lamp bead marks of the corresponding LED lamp beads, and historical light source parameters are generated by integrating them in the order of the lamp bead marks.

9. The remotely programmable IC dimming and color-adjusting RGB multi-color LED light source according to claim 8, characterized in that: The set trajectory includes a counterclockwise trajectory, a clockwise trajectory or a linear trajectory.

10. The remotely programmable IC dimming and color-adjusting RGB multi-color LED light source according to claim 1, characterized in that: The neural network 1 is a BP neural network model or a support vector machine model; the neural network 2 is a BP neural network model or a support vector machine model.

11. The remotely programmable IC dimming and color-adjusting RGB multi-color LED light source according to claim 1, characterized in that: The central control module is in communication connection with the data acquisition module; the data acquisition module acquires behavior data and environment data in the process of controlling the LED light source through data sensors.

12. The remotely programmable IC dimming and color-adjusting RGB multi-color LED light source according to claim 11, characterized in that: The data sensors include motion sensors, infrared sensors and light sensors.

Images