Light control method and system, electronic equipment and storage medium

Abstract

The invention provides a light control method and system, electronic equipment and a storage medium. The method comprises the steps that a user interaction request collected by Internet of Things equipment in real time is acquired; generating light scene configuration parameters matched with the user interaction request based on a pre-trained large language model; the large language model takes a standardized light scene in a preset scene library as a training sample, and learns corresponding light scene configuration parameters generated according to natural language description in a training process; the standardized lighting scene is jointly defined by natural language description, lighting scene configuration parameters and a mapping relation thereof; transmitting the generated light scene configuration parameters as effective loads of streaming data; and analyzing the lamp control instruction from the streaming data, and sending the lamp control instruction to the light equipment by adopting a sectional output strategy. According to the invention, dynamic and accurate light scene configuration is realized through natural language interaction, and the diversity of light atmosphere and the instantaneity of user experience are improved.

Description

Technical Field

[0001] The present application belongs to the field of lighting technology and relates to a lighting control method, system, electronic device and storage medium. Background Art

[0002] With the popularization of smart homes, lighting systems are gradually evolving from traditional single-switch control to scenario-based and personalized services. Users expect to control lights in a more natural and flexible way to meet their diverse needs in different scenarios such as rest, entertainment, and work.

[0003] In recent years, the vigorous development of artificial intelligence technology has opened up new technical paths for intelligent lighting control. However, existing solutions still have many limitations. For example, automated lighting control methods based on rule engines are limited by the size of the expert knowledge base and can only apply a limited number of standardized lighting scenarios, which cannot meet the diverse needs of users for complex scenarios. Although traditional machine learning methods have made some progress in predicting user behavior, they have the disadvantages of complex feature engineering and insufficient cross-scenario generalization capabilities, making it difficult to effectively understand and respond to abstract needs such as "atmosphere." Generative adversarial networks have been tried to be applied to the generation of light and shadow effects, but in actual applications they have problems such as slow response speed, unstable output, and poor semantic controllability. In addition, most existing lighting control methods adopt non-interactive designs, which make it difficult to dynamically adjust and optimize scenes based on immediate user feedback, resulting in a poor user experience. Summary of the Invention

[0004] The present application provides a lighting control method, system, electronic device and storage medium for providing personalized lighting services that meet the actual needs of users.

[0005] In a first aspect, the present application provides a lighting control method, comprising: obtaining user interaction requests collected in real time by an Internet of Things device; the user interaction requests are described in natural language; based on a pre-trained large language model, generating lighting scene configuration parameters that match the user interaction requests; the large language model uses standardized lighting scenes in a preset scene library as training samples, and learns to generate corresponding lighting scene configuration parameters based on natural language descriptions during training; the standardized lighting scene is defined by the natural language description, the lighting scene configuration parameters and their mapping relationship; the lighting scene configuration parameters include brightness, color temperature, hue and light effect; the generated lighting scene configuration parameters are transmitted as the payload of streaming data; lighting control instructions are parsed from the streaming data, and a segmented output strategy is adopted to send the lighting control instructions to the lighting equipment to control the lighting equipment to display the corresponding lighting scene.

[0006] In an implementation of the first aspect, generating lighting scene configuration parameters that match the user interaction request based on a pre-trained large language model includes: performing semantic parsing on the user interaction request to obtain a semantic parsing result; the semantic parsing result includes emotional tendency characteristics, color preference characteristics, and brightness requirement characteristics; performing similarity matching on the semantic parsing result with the standardized lighting scenes in the preset scene library; if the match is successful, calling the lighting scene configuration parameters corresponding to the standardized lighting scene; otherwise, generating new lighting scene configuration parameters based on a cross-domain knowledge graph; the knowledge elements in the cross-domain knowledge graph include color psychology, optical colorimetry, and real-time environmental perception information.

[0007] In one implementation of the first aspect, generating new lighting scene configuration parameters based on a cross-domain knowledge graph includes: generating a hue corresponding to the emotional tendency characteristic based on the association between color and emotion in color psychology; generating a color temperature corresponding to the color preference characteristic based on the blackbody radiation law in optical colorimetry; and dynamically generating a brightness corresponding to the brightness demand characteristic based on the real-time environmental perception information.

[0008] In an implementation of the first aspect, before transmitting the generated lighting scene configuration parameters as the payload of streaming data, the method further includes: performing format conversion on the lighting scene configuration parameters based on the large language model; wherein the step of performing format conversion on the lighting scene configuration parameters based on the large language model includes: determining the data type of the lighting scene configuration parameters; if the data type is scene data, buffering and regularizing the lighting scene configuration parameters; packaging the buffered and regularized lighting scene configuration parameters in blocks to obtain scene data blocks; if the data type is text data, directly transmitting the lighting scene configuration parameters in real time; and packaging the directly transmitted lighting scene configuration parameters in real time to obtain text data blocks.

[0009] In an implementation of the first aspect, a segmented output strategy is adopted to send the lighting control instruction to the lighting device, including: in a real-time response stage, the lighting control instruction is issued to the lighting device in real time based on a preset lighting control protocol, so that the lighting device can immediately complete the lighting state adjustment; in an asynchronous interpretation stage, a decision basis text is generated based on the large language model; the decision basis text is used to explain to the user the internal logic and design principles of the generation of the lighting scene configuration parameters.

[0010] In an implementation of the first aspect, it further includes: based on an incremental scene library update mechanism, storing the new lighting scene configuration parameters into the preset scene library; based on a continuous learning mechanism, dynamically optimizing the parameters of the large language model.

[0011] In one implementation of the first aspect, based on a continuous learning mechanism, dynamically optimizing the parameters of the large language model includes: collecting real-time feedback and evaluation of users on the lighting scene displayed by the lighting device; constructing a positive sample data set and a negative sample data set based on the feedback and evaluation; and adjusting the parameters of the large language model based on the constructed positive sample data set and the negative sample data set.

[0012] In the second aspect, the present application provides a lighting control system, including: an Internet of Things device for collecting user interaction requests in real time; a lighting control device, which is communicated with the Internet of Things device and is used to execute any of the methods described above; and a lighting device, which is communicated with the lighting control device and is used to display corresponding lighting scenes based on the lighting control instructions sent by the lighting control device.

[0013] In a third aspect, the present application provides an electronic device, comprising: a memory, the memory being used to store a computer program; and a processor, the processor being used to execute the computer program stored in the memory, so that the electronic device executes any one of the methods described above.

[0014] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which implements any of the methods described above when the computer program is executed by a processor.

[0015] As described above, the lighting control method, system, electronic device, and storage medium described in this application have the following beneficial effects:

[0016] (1) It fully utilizes the advantages of the large language model to accurately understand user intentions and generate personalized lighting scenes that meet the user's actual needs, thereby providing users with a comfortable, convenient and intelligent lighting experience;

[0017] (2) This application can directly call the lighting scene configuration parameters corresponding to the existing standardized lighting scenes without having to re-do the complex calculation and generation process, which greatly shortens the system response time. It can also generate new lighting scene configuration parameters based on the cross-domain knowledge graph, so that the system can cope with various non-standardized and complex application scenarios, providing more possibilities for generating new lighting scene configurations.

[0018] (3) By adopting a segmented output strategy that combines real-time response with asynchronous interpretation, it ensures rapid response to user requests and improves user trust;

[0019] (4) The cross-domain knowledge graph integrates color psychology, optical colorimetry, and real-time environmental perception information, which not only ensures that the generated lighting scene accurately meets user expectations, but also can flexibly respond to changes in the current environment, thus realizing intelligent lighting management. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 Shown is an application scenario architecture diagram of a lighting control method according to an embodiment of the present application.

[0021] Figure 2 Shown is a flow chart of a lighting control method according to an embodiment of the present application.

[0022] Figure 3 Shown is a diagram of the working principle of a large language model according to an embodiment of the present application.

[0023] Figure 4 Shown is a flow chart of format conversion according to an embodiment of the present application.

[0024] Figure 5 Shown is a flow chart of a lighting control method according to another embodiment of the present application.

[0025] Figure 6 Shown is a structural schematic diagram of a lighting control system according to an embodiment of the present application.

[0026] Figure 7 Shown is a structural schematic diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0027] The following describes the embodiments of the present application through specific examples. Those skilled in the art can easily understand the other advantages and effects of the present application from the content disclosed in this specification. The present application can also be implemented or applied through other different specific embodiments. The details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other unless they conflict.

[0028] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application. Therefore, the illustrations only show components related to the present application and are not drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component can be changed at will, and the component layout type may also be more complicated.

[0029] The following embodiments of this application provide a lighting control method, system, electronic device, and storage medium. This application implements dynamic and precise lighting scene configuration through natural language interaction, enhancing the diversity of lighting atmospheres and the immediacy of user experience.

[0030] See also Figure 1 , which shows an application scenario architecture diagram of the lighting control method according to an embodiment of the present application. Figure 1As shown, in this embodiment, user interaction requests are collected through IoT devices such as smartphones, smart bracelets, smart lamps, or smart speakers. These requests are semantically parsed using a large language model. Based on the semantic parsing results, standardized lighting scene configuration parameters are retrieved from a preset scene library, or new lighting scene configuration parameters are generated based on a cross-domain knowledge graph. These lighting scene configuration parameters are streamed to the lighting devices for display.

[0031] See also Figure 2 , which is a flow chart of a lighting control method according to an embodiment of the present application.

[0032] like Figure 2 As shown, this embodiment provides a lighting control method, including the following steps S100 to S400. The steps S100 to S400 can be executed in a local server or a cloud server.

[0033] In step S100, a user interaction request collected in real time by an IoT device is obtained; the user interaction request is described in natural language.

[0034] In this embodiment, IoT devices include, but are not limited to, smart speakers, smart lamps, smart central control panels, wearable devices, and smart appliances with integrated voice assistants. Specifically, an application (APP) designed specifically for lighting control can run on the IoT device, providing the user with an intuitive natural language interface. Through this natural language interface, users can conveniently interact with the IoT device using natural language.

[0035] User interaction requests are instructions containing multi-dimensional parameters, sent by users to IoT devices via a natural language interface. Specifically, user interaction requests include key information such as environmental characteristics, spatial positioning, and desired atmosphere. By semantically parsing user interaction requests, we can clearly understand the user's intent.

[0036] During actual interactions, the user's description may be irrelevant to the context. IoT devices can address this issue by providing intelligent reminders and dynamic guidance. For example, if the user's input is irrelevant to the context, the IoT device will promptly remind the user to enter the correct information. Furthermore, if the user enters ambiguous information, the IoT device will gradually guide the user to provide key information through multiple rounds of dialogue. This implementation not only effectively lowers the user's operational threshold but also significantly improves the error tolerance of interactions, making it easier and more convenient for users to complete interactions.

[0037] In actual applications, users can flexibly choose the interaction method based on their usage habits and actual needs. Natural language descriptions can include various forms such as voice and text. For example, users can directly communicate their needs to IoT devices through voice input, or manually enter text commands in the application's graphical interface. This diverse interaction method fully meets the needs of different users in different scenarios.

[0038] In step S200 , lighting scene configuration parameters matching the user interaction request are generated based on a pre-trained large language model.

[0039] Large Language Model (LLM) is an artificial intelligence technology based on deep learning.

[0040] In this embodiment, the large language model possesses mood sensing capabilities, keenly capturing the user's emotional state through voice and intonation analysis and contextual understanding of chat content. Furthermore, the large language model possesses identity awareness and can be combined with voiceprint recognition technology to provide differentiated services for multiple users. This allows accurate identification of different users and provides tailored lighting scene services based on their individual preferences and usage habits, meeting the unique needs of different users in various scenarios. Furthermore, the large language model possesses powerful semantic understanding capabilities, enabling an intelligent mapping between user expression, emotional needs, and lighting parameters. Regardless of the natural language used by the user to express their needs, the large language model can deeply understand the underlying emotional demands and accurately translate them into corresponding lighting parameter settings, enabling more precise and intelligent lighting control. The large language model also features streaming output capabilities. This means that the large language model can generate content in real time and continuously while processing user interaction requests, rather than waiting for the entire computational process to complete before outputting the results all at once. This feature enables more natural and fluid responses when interacting with users, providing users with immediate feedback and a more engaging experience.

[0041] The large language model used in this application can be Deepseek, Tongyi Qianwen or GPT-4. Other large language models with the same functions are also applicable to this application.

[0042] The large language model uses standardized lighting scenes in a preset scene library as training samples. During the training process, it learns to generate corresponding lighting scene configuration parameters based on natural language descriptions. The standardized lighting scenes are defined by natural language descriptions, lighting scene configuration parameters, and their mapping relationship. The lighting scene configuration parameters include brightness, color temperature, hue, and lighting effects.

[0043] This training method enables the large language model to fully understand the specific parameters of different scenarios and their applicable scenarios, so that when users actually interact, it can quickly generate matching lighting scene configuration parameters and provide users with better quality and more personalized services.

[0044] The preset scene library includes multiple standardized lighting scenes. Some of these scenes are shown in Table 1.

[0045] Table 1. Some standardized lighting scenes

[0046]

[0047] Taking the "Warm and Relaxing" scenario as an example, the corresponding natural language description for this scenario could be something like "I want to set a lighting scene that suits my home atmosphere," "Please configure a warm and relaxing lighting scene for me," and similar expressions. After training the large language model with different expressions, even when the user's input description is slightly distorted, the model can still accurately identify the intended scene as "warm and relaxing" and output the corresponding configuration parameters, namely "warm yellow light, 40% brightness, no flicker." The situation for other scenarios is similar and will not be repeated here.

[0048] It's important to note that the lighting scene configuration parameters generated by the large language model based on natural language descriptions include not only basic parameters such as brightness, color temperature, hue, and lighting effects, but may also include other related parameters such as light flickering frequency and gradient speed. The appropriate combination and setting of these parameters can further enrich the lighting scene effects and meet users' diverse lighting needs in different scenarios. For example, during special holidays or celebrations, setting a specific flickering frequency and gradient speed can create a cheerful and festive atmosphere.

[0049] This application fully utilizes the advantages of large language models, can accurately understand user intentions, and generate personalized lighting scenes that meet the user's actual needs, thereby bringing users a comfortable, convenient and intelligent lighting experience.

[0050] In step S300, the generated lighting scene configuration parameters are transmitted as a payload of streaming data.

[0051] By streaming lighting scene configuration parameters, this application allows lighting devices to immediately adjust lighting settings after receiving some of these parameters, without having to wait for all parameters to be generated and transmitted. This significantly shortens scene response time, allowing users to see lighting scene changes more quickly, ensuring a smooth user experience and enhancing user trust and trust.

[0052] In step S400, a lighting control instruction is parsed from the streaming data, and a segmented output strategy is adopted to send the lighting control instruction to the lighting device to control the lighting device to display a corresponding lighting scene.

[0053] In this embodiment, the segmented output strategy includes a real-time response phase and an asynchronous interpretation phase. This application adopts a segmented output strategy that combines real-time response with asynchronous interpretation, which not only ensures a quick response to user requests but also improves user trust.

[0054] When the lighting equipment receives the lighting control command, it will immediately start the corresponding processing flow and adjust the lighting settings in real time.

[0055] Specifically, the lighting equipment will accurately adjust parameters such as brightness, color temperature, hue, and lighting effects based on the lighting control instructions received to ensure that the lighting scene desired by the user can be presented quickly and accurately.

[0056] Traditional lighting configuration often requires manual adjustment of various parameters, which is not only inefficient but also prone to errors. This implementation uses automated lighting adjustment, significantly reducing the cost and workload of manual configuration, providing users with convenient and efficient services and improving user satisfaction.

[0057] The specific implementation principles and implementation methods of step S200 and step S400 of the lighting control method in this embodiment are described in detail below.

[0058] See also Figure 3 , which shows a working principle diagram of a large language model according to an embodiment of the present application.

[0059] like Figure 3 As shown, step S200 of generating lighting scene configuration parameters matching the user interaction request based on the pre-trained large language model includes the following steps S201 to S203.

[0060] In step S201, semantic parsing is performed on the user interaction request to obtain a semantic parsing result.

[0061] The large language model leverages its powerful mood perception, identity perception, and semantic understanding capabilities to not only capture the literal meaning of user input but also deeply analyze the underlying intent, emotional overtones, and context. During parsing, the large language model analyzes user input sentence by sentence and word by word, incorporating contextual information to eliminate ambiguity and ensure accurate understanding of every word and phrase. Furthermore, the large language model incorporates knowledge of common lighting scene configurations and user expression habits to further infer the user's true needs.

[0062] In one embodiment of the present application, the semantic analysis result includes emotional tendency features, color preference features, and brightness requirement features.

[0063] Specifically, emotional propensity features refer to the emotional state or emotional tendency expressed by users in their interaction requests. For example, when a user mentions words like "celebration" or "party," the large language model can infer that the user's emotional propensity is positive and upbeat. On the other hand, when a user mentions "rest" or "reading," it may indicate that the user is seeking to create a peaceful and comfortable atmosphere.

[0064] Color preference refers to a user's preference or inclination for colors in lighting scenarios. For example, a user might explicitly mention "I like blue lights," or imply a preference for a certain color by describing a scene like "a night at the seaside."

[0065] The brightness requirement feature refers to the user's specific needs or preferences for brightness in a lighting scenario. For example, a user may mention "needing bright light for work" or "preferring soft light for relaxation."

[0066] In step S202, similarity matching is performed between the semantic parsing result and the standardized lighting scenes in the preset scene library.

[0067] Specifically, the semantic parsing results and the standardized lighting scenes are first converted into comparable vectors. The cosine similarity between these vectors is then calculated to assess the degree of matching. Ultimately, the preset scene with the highest similarity is selected as the matching result. This implementation is not only computationally simple and efficient, but also effectively processes high-dimensional data, enabling rapid matching within a vast library of preset scenes.

[0068] In step S203, if the match is successful, the lighting scene configuration parameters corresponding to the standardized lighting scene are called; otherwise, new lighting scene configuration parameters are generated based on the cross-domain knowledge graph.

[0069] For example, when a user's interactive request is "warm dinner lighting," the large language model will perform semantic analysis, extract key features such as "warmth" and "dinner," and convert them into a feature vector. Next, the cosine similarity between this feature vector and the feature vectors corresponding to each standardized lighting scene in the preset scene library is calculated. Suppose there is a scene called "Warm Rest" in the preset scene library, whose configuration parameters include warm yellow light, 40% brightness, and no flicker, which are highly consistent with the features entered by the user. In this case, the cosine similarity between the feature vector of the "Warm Rest" scene and the feature vector entered by the user will be very high, indicating that the two are highly similar. Therefore, the "Warm Rest" scene will be recommended to the user as a matching result, thus satisfying the user's need for warm dinner lighting.

[0070] In this implementation, when a match is successful, the lighting scene configuration parameters corresponding to the existing standardized lighting scene are directly called, eliminating the need for complex recalculation and generation. This significantly reduces system response time, allows for rapid lighting scene setup, and improves the user experience. Furthermore, the configuration parameters for standardized lighting scenes are typically carefully designed and tested to ensure optimal lighting effects and performance in the corresponding scenario. Directly calling these parameters avoids the potential uncertainty and errors associated with regeneration, ensuring the stability and consistency of the lighting scene.

[0071] The knowledge elements in the cross-domain knowledge graph include color psychology, optical colorimetry and real-time environmental perception information.

[0072] In one embodiment of the present application, step S204 of generating new lighting scene configuration parameters based on the cross-domain knowledge graph may include: generating the hue corresponding to the emotional tendency characteristic based on the association between color and emotion in the color psychology; generating the color temperature corresponding to the color preference characteristic based on the blackbody radiation law in the optical colorimetry; and dynamically generating the brightness corresponding to the brightness demand characteristic based on the real-time environmental perception information.

[0073] In this embodiment, the relationship between color and emotion in color psychology and the blackbody radiation law in optical colorimetry are both existing technologies and will not be discussed in detail here. Real-time environmental perception information includes parameters such as the current light intensity, temperature, and humidity, acquired by photosensors. This application can dynamically adjust lighting based on these parameters to ensure that the lighting is always adapted to the current environment.

[0074] For example, during the day when there is plenty of sunlight, the brightness of the light will be reduced to avoid discomfort caused by excessive light; while at night or in a dark environment, the brightness of the light will be increased to ensure sufficient lighting effect. In addition, this application will further fine-tune the brightness of the light according to the user's activity status (such as reading, resting, exercising, etc.) to provide a more comfortable and healthy lighting environment.

[0075] In this implementation, new lighting scene configuration parameters are generated based on a cross-domain knowledge graph, enabling the system to handle a variety of non-standardized and complex application scenarios, providing more possibilities for generating new lighting scene configurations. Furthermore, the cross-domain knowledge graph integrates information from color psychology, optical colorimetry, and real-time environmental perception. This ensures that the generated lighting scenes precisely meet user expectations and flexibly respond to changes in the current environment, enabling intelligent lighting management.

[0076] In one embodiment of the present application, before transmitting the generated lighting scene configuration parameters as a payload of streaming data, the method further includes: performing format conversion on the lighting scene configuration parameters based on the large language model.

[0077] See also Figure 4 , which is a flowchart showing format conversion according to an embodiment of the present application.

[0078] like Figure 4 As shown, the step of converting the format of the lighting scene configuration parameters based on the large language model includes: determining the data type of the lighting scene configuration parameters; if the data type is scene data, buffering and regularizing the lighting scene configuration parameters; packaging the buffered and regularized lighting scene configuration parameters in blocks to obtain scene data blocks; if the data type is text data, directly transmitting the lighting scene configuration parameters in real time; packaging the light scene configuration parameters that have been directly transmitted in real time in blocks to obtain text data blocks.

[0079] In this embodiment, scene data refers to the parameters and settings associated with a specific lighting scene. This data describes the specific configuration parameters of the lighting scene. Text data refers to information in text form, such as descriptions and annotations, that may be relevant to the use of the scene.

[0080] In this implementation, the lighting scene configuration parameters are transmitted in data blocks, which can ensure the integrity and regularity of the data, reduce the pressure on the system during data transmission and processing, improve the stability and reliability of the system, and facilitate parsing and processing at the receiving end.

[0081] See also Figure 5 , which is a flow chart of a lighting control method according to another embodiment of the present application.

[0082] In an embodiment of the present application, the segmented output strategy described in step S400 to send the lighting control instruction to the lighting device may include the following steps S401 and S402.

[0083] In step S401, in the real-time response stage, the lighting control instruction is sent to the lighting device in real time based on the preset lighting control protocol, so that the lighting device can immediately complete the lighting status adjustment.

[0084] In this embodiment, the preset lighting control protocol adopts a lightweight design and can complete the issuance of instructions within 50ms, fully ensuring immediacy.

[0085] The real-time response stage can quickly transmit key basic lighting parameters, ensure that the lighting equipment responds immediately, and realize the initial switching of scenes, so that users can see the effects in a short time, meet the needs of instant feedback, and improve the smoothness and satisfaction of the experience.

[0086] In step S402, during the asynchronous interpretation phase, a decision basis text is generated based on the large language model; the decision basis text is used to explain to the user the inherent logic and design principles of the generation of the lighting scene configuration parameters.

[0087] The asynchronous explanation stage generates decision-making basis text, which can help users understand the logic and reasons for scenario generation, resolve their confusion, and enhance trust.

[0088] In one embodiment of the present application, the lighting control method described in this embodiment further includes step S500 and step S600.

[0089] In step S500, based on the incremental scene library update mechanism, the new lighting scene configuration parameters are stored in the preset scene library.

[0090] Preset scenario libraries usually have problems such as insufficient flexibility, inefficient manual configuration, lack of personalized adaptation capabilities, and fragmentation of multimodal data.

[0091] The incremental scene library update mechanism in this embodiment enriches scene coverage by storing new lighting scene configuration parameters in the scene library, meeting users' diverse and personalized scene needs. Furthermore, the scene library can be automatically updated without manual configuration, reducing maintenance costs and improving update efficiency. Furthermore, the incremental update mechanism effectively integrates multimodal data such as environmental perception and user preferences, enhancing the accuracy and adaptability of scene generation.

[0092] In step S600 , the parameters of the large language model are dynamically optimized based on a continuous learning mechanism.

[0093] In one embodiment of the present application, based on a continuous learning mechanism, dynamically optimizing the parameters of the large language model includes: collecting real-time user feedback and evaluations of the lighting scene displayed by the lighting device; constructing a positive sample data set and a negative sample data set based on the feedback and evaluation; and adjusting the parameters of the large language model based on the constructed positive sample data set and the negative sample data set.

[0094] In this implementation, the parameters of the large language model can be continuously optimized through a continuous learning mechanism, enabling the model to more accurately understand user intent, generate scenario configurations that are more in line with personalized needs, and improve the accuracy and satisfaction of scenario recommendations.

[0095] The following will provide additional explanation of the technical solution of this application in conjunction with two specific embodiments.

[0096] Case 1:

[0097] The user interaction request is: "It's Chinese New Year now, and my friends will come to my house to play later."

[0098] Based on the user's interaction request, the large language model matches the preset scene "Fireworks Year After Year." "Fireworks Year After Year" uses bright, festive warm tones to create a festive and lively home environment, perfect for Lunar New Year celebrations and gatherings with friends. The brightness is set to 91%, ensuring the room is vibrant without being overly glaring. The color scheme is based on color psychology, with red and orange tones evoking Spring Festival elements such as fireworks and lanterns, inspiring joy. The warm colors also enhance the warmth of social interactions.

[0099] Specifically, the configuration parameters corresponding to "Fireworks Year After Year" are:

[0100] {"id":"5-1","brightness":91,"colorArr":[{"x":45172,"y":19909},{"x":42651,"y":22070} ,{"x":38600,"y":23883},{"x":37022,"y":25236},{"x":44925,"y":19388}],"colorMode":1};

[0101] Multiple sets of xy represent multiple sets of color data. It should be noted that this JSON data will not be displayed on the natural language interaction interface of the application.

[0102] Case 2:

[0103] The user interaction request is: "The Yellow River flows into the sea."

[0104] "The Yellow River Flows into the Sea" depicts a magnificent natural scene, so the lighting design needed to reflect a sense of fluidity and warmth. Based on user interaction requests, the large language model selects 80% brightness to convey vibrancy and clarity while avoiding glare. A gradient warm color scheme is used, transitioning from bright gold (FFD79D) to deep orange (FF963E), emulating the flow of a river illuminated by the setting sun. This color scheme is based on color psychology. Warm tones convey a sense of warmth and comfort, while the gradient effect enhances the dynamic beauty of the space, creating a poetic and relaxing environment.

[0105] Specifically, the configuration parameters corresponding to "Yellow River Flow into the Sea" are:

[0106] {"brightness":60,"colorArr":[{"x":29823,"y":26660},{"x":34175,"y":27008},{"x" :36768,"y":26323},{"x":39095,"y":25196},{"x":40939,"y":23830}],"colorMode":1}.

[0107] It should be noted that the protection scope of the lighting control method described in the embodiment of the present application is not limited to the execution order of the steps listed in this embodiment. All solutions implemented by adding, reducing, or replacing steps in the existing technology based on the principles of the present application are included in the protection scope of the present application.

[0108] See also Figure 6 , which is a structural diagram of a lighting control system according to an embodiment of the present application.

[0109] like Figure 6 As shown, this embodiment provides a lighting control system, including an Internet of Things device, a lighting control device and lighting equipment.

[0110] The IoT device is used to collect user interaction requests in real time.

[0111] The lighting control device is communicatively connected to the Internet of Things device and is used to execute any of the lighting control methods described above.

[0112] Specifically, the light control device includes an application program for executing the light control method.

[0113] In other embodiments, the application program for executing the lighting control method may also be in any networked device.

[0114] The lighting equipment is in communication with the lighting control device and is configured to display corresponding lighting scenes based on the lighting control instructions sent by the lighting control device.

[0115] It should be noted that the embodiments related to IoT devices and lighting devices have been described in detail in the above-mentioned lighting control method and will not be repeated here. The structure and principle of the lighting control device correspond to the steps in the above-mentioned lighting control method and will not be repeated here.

[0116] See also Figure 7 , which is a schematic structural diagram of an electronic device according to an embodiment of the present application.

[0117] like Figure 7 As shown, this embodiment provides an electronic device including a memory and a processor.

[0118] The memory is used to store computer programs.

[0119] The processor is configured to execute the computer program stored in the memory, so as to enable the electronic device to execute any one of the above-mentioned light control methods.

[0120] It will be appreciated that the memory may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM) or a programmable read-only memory (PROM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM) and synchronous static random access memory (SSRAM). The memory described in the embodiments of the present invention is intended to include, but is not limited to, these and any other suitable categories of memory.

[0121] The memory in the embodiments of the present application is used to store various types of data to support the operation of the electronic device. Examples of this data include any executable program used to operate on the electronic device, such as an operating system and application programs. An operating system includes various system programs, such as a framework layer, a core library layer, and a driver layer, which are used to implement various basic services and handle hardware-based tasks. The lighting control method provided in the embodiments of the present invention can be included in an application program.

[0122] The method disclosed in the above-mentioned embodiment of the present application can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During the implementation process, each step of the above method can be completed by an integrated logic circuit of the hardware in the processor or an instruction in the form of software. The above-mentioned processor may be a general-purpose processor, a digital signal processor (DSP, Digital Signal Processor), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor can be a microprocessor or any conventional processor, etc. The steps of the accessory optimization method provided in conjunction with the embodiment of the present invention can be directly embodied as being executed by a hardware decoding processor, or being executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium, which is located in a memory, and the processor reads the information in the memory and completes the steps of the above-mentioned method in combination with its hardware.

[0123] This embodiment also includes at least one network interface and a user interface. The various components in the device are coupled together via a bus system. It will be appreciated that the bus system is used to enable communication between these components. In addition to a data bus, the bus system also includes a power bus, a control bus, and a status signal bus. However, for clarity, each bus is labeled as a bus system in Figure X. The user interface may include a display, keyboard, mouse, trackball, click gun, keys, buttons, touchpad, or touch screen.

[0124] In an exemplary embodiment, the electronic device may be used by one or more application specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), and complex programmable logic devices (CPLDs) to perform the aforementioned method.

[0125] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices or methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of modules / units is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or units can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules or units, which can be electrical, mechanical or other forms.

[0126] The modules / units described as separate components may or may not be physically separate, and the components displayed as modules / units may or may not be physical modules, that is, they may be located in one place or distributed across multiple network elements. Some or all of the modules / units may be selected according to actual needs to achieve the purpose of the embodiments of the present application. For example, the functional modules / units in the various embodiments of the present application may be integrated into a processing module, or each module / unit may exist physically separately, or two or more modules / units may be integrated into a single module / unit.

[0127] Those skilled in the art should further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the composition and steps of each example according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0128] The present application also provides a computer-readable storage medium having a computer program stored thereon, and the computer program implements any of the above-described methods when executed by a processor. A person of ordinary skill in the art will appreciate that all or part of the steps in the method for implementing the above-described embodiment can be completed by instructing the processor through a program, and the program can be stored in a computer-readable storage medium, and the storage medium is a non-transitory medium, such as a random access memory, a read-only memory, a flash memory, a hard disk, a solid-state drive, a magnetic tape, a floppy disk, an optical disc, and any combination thereof. The above-mentioned storage medium can be any available medium that a computer can access or a data storage device such as a server or a data center that includes one or more available media. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a digital video disc (DVD)), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0129] The descriptions of the processes or structures corresponding to the above figures have different emphases. For parts that are not described in detail in a certain process or structure, please refer to the relevant descriptions of other processes or structures.

[0130] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Anyone skilled in the art may modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by one of ordinary skill in the art without departing from the spirit and technical concepts disclosed in this application shall be covered by the claims of this application.

Claims

1. A lighting control method, characterized in that: include: Obtain user interaction requests collected in real time by IoT devices; The user interaction request is described in natural language; Generate lighting scene configuration parameters that match the user interaction request based on a pre-trained large language model; The large language model uses standardized lighting scenes in a preset scene library as training samples, and learns to generate corresponding lighting scene configuration parameters based on natural language descriptions during the training process; the standardized lighting scenes are defined by the natural language descriptions, the lighting scene configuration parameters, and their mapping relationship; The lighting scene configuration parameters include brightness, color temperature, hue and lighting effect; Transmitting the generated lighting scene configuration parameters as a payload of streaming data; A lighting control instruction is parsed from the streaming data, and a segmented output strategy is adopted to send the lighting control instruction to the lighting device to control the lighting device to display a corresponding lighting scene.

2. The method according to claim 1, characterized in that Generating lighting scene configuration parameters matching the user interaction request based on the pre-trained large language model includes: Performing semantic analysis on the user interaction request to obtain a semantic analysis result; the semantic analysis result includes an emotional tendency feature, a color preference feature, and a brightness requirement feature; Performing similarity matching between the semantic parsing result and the standardized lighting scenes in the preset scene library; If the match is successful, the lighting scene configuration parameters corresponding to the standardized lighting scene are called; Otherwise, new lighting scene configuration parameters are generated based on a cross-domain knowledge graph; the knowledge elements in the cross-domain knowledge graph include color psychology, optical colorimetry, and real-time environmental perception information.

3. The method according to claim 2, characterized in that Based on the cross-domain knowledge graph, new lighting scene configuration parameters are generated, including: Based on the association between color and emotion in color psychology, generating a color tone corresponding to the emotional tendency feature; generating a color temperature corresponding to the color preference characteristic based on the blackbody radiation law in optical colorimetry; Based on the real-time environmental perception information, a brightness corresponding to the brightness requirement feature is dynamically generated.

4. The method according to claim 1, wherein Before transmitting the generated lighting scene configuration parameters as a payload of streaming data, the method further includes: performing format conversion on the lighting scene configuration parameters based on the large language model; wherein the step of converting the format of the lighting scene configuration parameters based on the large language model includes: Determining the data type of the lighting scene configuration parameter; If the data type is scene data, buffering and regularizing the lighting scene configuration parameters; and packaging the buffered and regularized lighting scene configuration parameters into blocks to obtain scene data blocks. If the data type is text data, the lighting scene configuration parameters are directly transmitted in real time; and the lighting scene configuration parameters directly transmitted in real time are divided into blocks and packaged to obtain text data blocks.

5. The method according to claim 1, wherein Using a segmented output strategy to send the lighting control command to the lighting device includes: In the real-time response stage, the lighting control instruction is issued to the lighting device in real time based on the preset lighting control protocol, so that the lighting device can immediately complete the lighting status adjustment; During the asynchronous interpretation stage, a decision-making basis text is generated based on the large language model; the decision-making basis text is used to explain to the user the inherent logic and design principles of the generation of the lighting scene configuration parameters.

6. The method according to claim 2, characterized in that Also includes: Based on the incremental scene library update mechanism, the new lighting scene configuration parameters are stored in the preset scene library; Based on a continuous learning mechanism, the parameters of the large language model are dynamically optimized.

7. The method according to claim 6, characterized in that Based on the continuous learning mechanism, the parameters of the large language model are dynamically optimized including: Collecting user feedback and evaluation on the lighting scene displayed by the lighting device in real time; Based on the feedback evaluation, construct a positive sample data set and a negative sample data set; Based on the constructed positive sample dataset and the negative sample dataset, the parameters of the large language model are adjusted.

8. A lighting control system, characterized in that: include: IoT devices, used to collect user interaction requests in real time; A lighting control device, communicatively connected to the IoT device, and configured to execute the method according to any one of claims 1 to 7; The lighting equipment is in communication with the lighting control device and is used to display corresponding lighting scenes based on the lighting control instructions sent by the lighting control device.

9. An electronic device, characterized in that: include: a memory for storing a computer program; A processor, wherein the processor is configured to execute the computer program stored in the memory so as to enable the electronic device to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images