A smart voice-controlled indoor lighting system

Abstract

This invention provides an intelligent voice-controlled indoor lighting system, comprising: a voice recognition module for acquiring and analyzing indoor sound information to obtain a first analysis result; an auxiliary recognition module for acquiring and analyzing indoor image data, electrical equipment data, and sensor data to obtain a second analysis result; a central control module for determining corresponding lighting control commands based on the first and second analysis results; and a lighting control module for controlling the indoor lighting equipment according to the lighting control commands. This invention utilizes various types of auxiliary recognition data to provide a basis for voice control, enhancing the accuracy of user intent recognition during voice control and improving the recognition capability of the voice-controlled lighting system.

Description

Technical Field

[0001] This invention relates to the field of intelligent lighting control technology, and in particular to an intelligent voice-controlled indoor lighting system. Background Technology

[0002] Lighting is an indispensable part of modern human life and work. According to our usage habits, traditional home lighting switches are almost entirely manual buttons. This has long limited people's imagination for home lighting design, making them inconvenient to use and posing certain safety hazards. Even if remote controls are available, they are mostly infrared. Dimming is also mostly manual physical dimming, lacking energy-saving and memory functions. Furthermore, their protective effect on the eyes is limited.

[0003] Compared to traditional lighting, smart lighting enables fully automatic dimming. For example, a smart lighting system can be set to several basic states, which automatically adjust to each other according to preset times, intelligently adjusting the illuminance to the most suitable level. After installing a smart lighting system, natural light sources are fully utilized. Natural light can be adjusted and controlled via smart devices with light-control functions, and it can also be linked with the lighting system. When the weather changes, the system will automatically adjust to maintain the stability of indoor lighting. Smart lighting also extends the lifespan of light sources. A smart lighting system can intelligently dim most light fixtures, providing sufficient lighting in the required spaces and at the required times, and can promptly turn off unnecessary lights, making full use of natural light, thereby reducing energy consumption and easing the burden on the power supply for the homeowner. The main cause of light source damage is overvoltage in the power grid; simply reducing the operating voltage can extend the lifespan of the light source. The smart lighting control system uses a soft-start method, which can control the power grid's inrush voltage and surge voltage, protecting the filament from thermal shock and thus extending its lifespan.

[0004] Therefore, intelligent lighting is an important direction for the development of social sciences. Existing indoor voice-controlled intelligent lighting systems still adopt a relatively simple control method, namely the direct recognition mode. When using the voice control function, users often make incorrect judgments due to the overly simplistic voice control judgment mechanism. This directly leads to a poor user experience with this type of voice-controlled intelligent lighting system, which seriously hinders the popularization of intelligent lighting systems. Summary of the Invention

[0005] This invention provides an intelligent voice-controlled indoor lighting system that can utilize various types of auxiliary recognition data to provide judgment criteria for voice control, enhance the accuracy of user intent recognition during the voice control process, and improve the recognition capability of the voice-controlled lighting system.

[0006] This invention provides an intelligent voice-controlled indoor lighting system, comprising:

[0007] The sound recognition module is used to acquire sound information present indoors and analyze the sound information to obtain the first analysis result;

[0008] The auxiliary recognition module is used to acquire indoor image data, electrical equipment data, and sensor data, and analyze them to obtain a second analysis result.

[0009] The central control module is used to determine the corresponding lighting control commands based on the first analysis result and the second analysis result;

[0010] The lighting control module is used to control indoor lighting equipment according to lighting control commands.

[0011] Preferably, the voice recognition module includes:

[0012] The sound acquisition unit is used to acquire sound information present indoors;

[0013] The voice extraction unit is used to extract voice data from the acquired sound information.

[0014] The speech recognition unit is used to analyze human voice data to obtain semantic text through a preset speech recognition model; wherein, the speech recognition model is trained in advance using the user's speech training file;

[0015] The keyword recognition unit is used to extract keywords from semantic text based on a preset keyword library and obtain the extraction results;

[0016] The key sound extraction unit is used to identify key sounds in sound information based on a preset key sound frequency feature library to obtain the identification result;

[0017] The first result generation unit is used to obtain the first analysis result based on the extraction result and the recognition result.

[0018] Preferably, the auxiliary recognition module includes:

[0019] The video acquisition unit is used to acquire video information appearing indoors;

[0020] The video analysis unit is used to analyze video information, determine the human posture and gesture features, spatial relationship information of home appliances and lamps in the video, and generate video analysis results.

[0021] The electrical analysis unit is used to acquire the operating status information of indoor electrical equipment and analyze the operating status to obtain electrical analysis results;

[0022] The sensing analysis unit is used to acquire sensing data from indoor sensing devices and analyze the sensing data to obtain sensing data analysis results.

[0023] The second result generation unit is used to generate a second analysis result based on the video analysis result, the electrical analysis result, and the sensor data analysis result.

[0024] Preferably, the central control module includes:

[0025] The instruction selection unit is used to select the corresponding lighting control instruction from the first analysis result and the second analysis result, and to send the lighting control instruction to the lighting control module.

[0026] The path prediction unit is used to determine the user's location indoors based on the first analysis result and the second analysis result, predict the user's movement path, obtain the prediction result, and send it to the lighting control module as a control reference.

[0027] The model training unit is used to create a speech recognition model, and at the same time, it acquires the user's speech data and uses it as a speech training file to train the speech recognition model.

[0028] Preferably, the instruction selection unit performs the following:

[0029] Obtain all first-type instructions and second-type instructions from the preset candidate instruction set of each lighting functional unit. The first-type instructions are the control instructions set for each basic element in the first analysis result, and the second-type instructions are the dependency instructions set for each basic element in the second analysis result.

[0030] Based on the results of the first analysis and the second analysis, multiple heuristic quantities corresponding to each first type of instruction in each lighting functional unit are determined. The heuristic quantities include the dependency quantity between the first type of instruction and the second type of instruction, the relationship quantity between the first type of instruction and the processing unit, and the relationship quantity between the first type of instruction and the lighting functional unit.

[0031] The first type of instructions in each lighting functional unit are sorted multiple times. During each sorting, one of the heuristic quantities is selected as the sorting comparison quantity according to the priority order of each heuristic quantity to obtain the sorted first type of instruction sequence and generate lighting control instructions.

[0032] Preferably, the first type of instruction in each lighting functional unit performs multiple sorting steps, including:

[0033] Step S1: Establish a first sorting sequence using all first-type instructions in each lighting functional unit, and select the heuristic with the highest priority as the first heuristic.

[0034] Step S2: Sort the first type of instructions a second time according to the value of the first heuristic, and divide the second sorted first type of instructions into multiple instruction subsequences according to the value of the heuristic, to obtain multiple sorted instruction subsequences;

[0035] Step S3: Take the optimal instruction subsequence corresponding to the optimal heuristic among the multiple sorted instruction subsequences as the first sorting sequence, and select the heuristic with the next priority as the first heuristic. Return to step S2 and continue until the sorting based on all heuristics is completed.

[0036] Preferably, the path prediction unit performs the following steps:

[0037] Based on the results of the first analysis, the keywords are determined, and based on the pre-defined mapping relationship between the keywords and each geographical location in the room, the target location corresponding to each keyword is determined.

[0038] The target location corresponding to each keyword is determined by statistical analysis, and the target location that appears most frequently is selected as the target location for movement.

[0039] Based on the preset room layout, determine the path connection between the user's current location and the target location, and use it as the prediction result of the user's movement path;

[0040] The prediction results are sent to the lighting control module as a control reference.

[0041] Preferably, the model training unit includes:

[0042] The feature training subunit is used to acquire the voice data of the target user in daily communication and build the user's voice dataset. The voice dataset is then input into a general speech recognition model to train and extract the user's voice features.

[0043] The speech enhancement subunit is used to acquire the speech data of the target user reading a specific text, and to enhance the speech data by using the user's speech characteristics to obtain enhanced speech data.

[0044] The model training subunit is used to create a speech recognition model based on a preset keyword library and train it using enhanced speech data to obtain the trained speech recognition model.

[0045] Preferably, the model training unit further includes:

[0046] The scenario creation subunit is used to establish the acoustic environment background for a specific speech recognition scenario and to obtain the background sound information under that acoustic environment background;

[0047] The speech synthesis subunit is used to synthesize synthesized speech data by using enhanced speech data and background sound information;

[0048] The context model training subunit is used to train a speech recognition model targeting a keyword library using synthesized speech data as the target input. After training, it establishes a mapping relationship between the model and the background sound environment to obtain a speech recognition model under that background sound environment.

[0049] Preferably, the lighting control module includes:

[0050] The instruction receiving unit is used to receive lighting control instructions issued by the central control module;

[0051] The instruction analysis unit is used to analyze lighting control instructions, determine the equipment control channels of the lighting equipment that need to be controlled, and the control signals of the lighting equipment obtained from the analysis.

[0052] The instruction execution unit is used to control the relevant lighting equipment according to the equipment control channel and the control signals of the lighting equipment obtained by parsing.

[0053] The present invention achieves the following beneficial effects:

[0054] 1. It can utilize various types of auxiliary recognition data to provide judgment basis for voice control, enhance the accuracy of user intent recognition during the voice control process, and improve the recognition capability of voice-controlled lighting systems.

[0055] 2. It realizes the determination of lighting control commands by using multiple data from the first and second analysis results, realizes the prediction and verification of user intentions, and determines the corresponding lighting control commands to control the lighting equipment based on the prediction and verification results.

[0056] 3. By extracting the user's speech features, the user's recitation speech data is enhanced to obtain enhanced speech data, highlighting the speech characteristics of the user in the recitation of specific texts. Then, a speech recognition model is created based on a preset keyword library and trained using the enhanced speech data. The trained speech recognition model has good adaptability to user semantic recognition, can obtain more accurate recognition results, and reduce the occurrence of misrecognition.

[0057] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings.

[0058] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0059] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0060] Figure 1 This is a schematic diagram of the structure of an intelligent voice-controlled indoor lighting system according to an embodiment of the present invention;

[0061] Figure 2 This is a schematic diagram of the central control module in an embodiment of the present invention;

[0062] Figure 3 This is a flowchart illustrating the steps of executing multiple sorting operations for the first type of instruction in an embodiment of the present invention. Detailed Implementation

[0063] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0064] The present invention provides an intelligent voice-controlled indoor lighting system, referring to... Figure 1 ,include:

[0065] The sound recognition module is used to acquire sound information present indoors and analyze the sound information to obtain the first analysis result;

[0066] The auxiliary recognition module is used to acquire indoor image data, electrical equipment data, and sensor data, and analyze them to obtain a second analysis result.

[0067] The central control module is used to determine the corresponding lighting control commands based on the first analysis result and the second analysis result;

[0068] The lighting control module is used to control indoor lighting equipment according to lighting control commands.

[0069] The working principle and beneficial effects of the above technical solution are as follows: A voice recognition module acquires indoor sound information and analyzes it to obtain a first analysis result; an auxiliary recognition module acquires and analyzes indoor image data, electrical equipment data, and sensor data to obtain a second analysis result; a central control module determines the corresponding lighting control command based on the first and second analysis results, thus using multiple analysis results as a basis for judging user intent to determine appropriate lighting control commands; finally, a lighting control module uniformly controls the indoor lighting equipment according to the lighting control commands. This utilizes multiple types of auxiliary recognition data to provide a basis for judgment in voice control, enhancing the accuracy of user intent recognition during the voice control process and improving the recognition capability of the voice-controlled lighting system.

[0070] In a preferred embodiment, the voice recognition module includes:

[0071] The sound acquisition unit is used to acquire sound information present indoors;

[0072] The voice extraction unit is used to extract voice data from the acquired sound information.

[0073] The speech recognition unit is used to analyze human voice data to obtain semantic text through a preset speech recognition model; wherein, the speech recognition model is trained in advance using the user's speech training file;

[0074] The keyword recognition unit is used to extract keywords from semantic text based on a preset keyword library and obtain the extraction results;

[0075] The key sound extraction unit is used to identify key sounds in sound information based on a preset key sound frequency feature library to obtain the identification result;

[0076] The first result generation unit is used to obtain the first analysis result based on the extraction result and the recognition result.

[0077] The working principle and beneficial effects of the above technical solution are as follows: The sound acquisition unit acquires indoor sound information; the voice extraction unit extracts voice data from the acquired sound information; the speech recognition unit analyzes the voice data using a preset speech recognition model to obtain semantic text; wherein, the speech recognition model is pre-trained using the user's voice training file; the keyword recognition unit extracts keywords from the semantic text based on a preset keyword library to obtain extraction results; the key sound extraction unit identifies key sounds from the sound information based on a preset key audio frequency feature library to obtain recognition results. The identified key sounds can be used to predict the user's intent, providing a reference for controlling lighting equipment. For example, the sound of a doorbell or the sound of a rice cooker finishing cooking can be used to predict the user's action intent, thus facilitating the control of lighting equipment along the user's path to the door or rice cooker; finally, the first result generation unit obtains the first analysis result based on the extraction and recognition results. This enables the acquisition, processing, and semantic analysis of indoor sound information. It extracts keywords from the semantic text according to a preset keyword library for determining control commands, accurately obtaining the user's voice recognition results. Furthermore, the key sound extraction unit identifies key sounds from the sound information based on a preset key audio frequency feature library, providing a reference for predicting the user's behavioral intentions.

[0078] In a preferred embodiment, the auxiliary identification module includes:

[0079] The video acquisition unit is used to acquire video information appearing indoors;

[0080] The video analysis unit is used to analyze video information, determine the human posture and gesture features, spatial relationship information of home appliances and lamps in the video, and generate video analysis results.

[0081] The electrical analysis unit is used to acquire the operating status information of indoor electrical equipment and analyze the operating status to obtain electrical analysis results;

[0082] The sensing analysis unit is used to acquire sensing data from indoor sensing devices and analyze the sensing data to obtain sensing data analysis results.

[0083] The second result generation unit is used to generate a second analysis result based on the video analysis result, the electrical analysis result, and the sensor data analysis result.

[0084] The working principle and beneficial effects of the above technical solution are as follows: The video acquisition unit acquires video information appearing indoors; the video analysis unit analyzes the video information to determine the human posture and gesture features, spatial relationship information of home appliances and lighting fixtures, and generates video analysis results. The human posture and gesture features can support the user's intention expressed through voice. For example, if a user lies in bed and says they want to go to the toilet while simultaneously getting up, the system has reason to believe the user intends to go to the toilet and can control the lighting equipment accordingly, turning on the lights along the path to the toilet. The electrical analysis unit acquires the operating status information of indoor electrical equipment and analyzes the operating status to obtain electrical analysis results. The spatial relationship information of home appliances and lighting fixtures allows the system to... Determining the distribution of available spaces in the indoor environment facilitates better prediction of user movement paths and verification of user behavior. For example, if a user mentions a rice cooker during a conversation while it is running, it's reasonable to assume the user intends to check it. Similarly, when a user is showering, the intensity of the bathroom heater light can be determined based on the shower noise, the water heater's operating status, and the external temperature. The sensor analysis unit acquires and analyzes sensor data from indoor devices to obtain analysis results. This data provides a reference for controlling, for example, the bathroom heater light or the ambient light. The second result generation unit generates a second analysis result based on video analysis, electrical analysis, and sensor data analysis results. This enables the collection and analysis of auxiliary recognition data, thereby improving the accuracy of user intent recognition through semantic analysis.

[0085] In a preferred embodiment, refer to Figure 2 The central control module includes:

[0086] The instruction selection unit is used to select the corresponding lighting control instruction from the first analysis result and the second analysis result, and to send the lighting control instruction to the lighting control module.

[0087] The path prediction unit is used to determine the user's location indoors based on the first analysis result and the second analysis result, predict the user's movement path, obtain the prediction result, and send it to the lighting control module as a control reference.

[0088] The model training unit is used to create a speech recognition model, and at the same time, it acquires the user's speech data and uses it as a speech training file to train the speech recognition model.

[0089] The working principle and beneficial effects of the above technical solution are as follows: The instruction selection unit selects the corresponding lighting control instruction based on the first and second analysis results, and sends the lighting control instruction to the lighting control module; the path prediction unit determines the user's indoor location based on the first and second analysis results and predicts the user's movement path, obtaining a prediction result, which is then sent to the lighting control module as a control reference; the model training unit creates a speech recognition model, simultaneously acquiring the user's speech data and using it as a speech training file to train the speech recognition model. This achieves decision analysis of lighting control instructions, prediction of user movement paths, and training of the speech recognition model through the central control module.

[0090] In a preferred embodiment, the instruction selection unit performs the following:

[0091] Obtain all first-type instructions and second-type instructions from the preset candidate instruction set of each lighting functional unit. The first-type instructions are the control instructions set for each basic element in the first analysis result, and the second-type instructions are the dependency instructions set for each basic element in the second analysis result.

[0092] Based on the results of the first analysis and the second analysis, multiple heuristic quantities corresponding to each first type of instruction in each lighting functional unit are determined. The heuristic quantities include the dependency quantity between the first type of instruction and the second type of instruction, the relationship quantity between the first type of instruction and the processing unit, and the relationship quantity between the first type of instruction and the lighting functional unit.

[0093] The first type of instructions in each lighting functional unit are sorted multiple times. During each sorting, one of the heuristic quantities is selected as the sorting comparison quantity according to the priority order of each heuristic quantity to obtain the sorted first type of instruction sequence and generate lighting control instructions.

[0094] The working principle and beneficial effects of the above technical solution are as follows: When selecting instructions, all first-type instructions and second-type instructions are obtained from the preset candidate instruction set of each lighting function unit. The first-type instructions are the control instructions set for each basic element in the first analysis result, such as turning on the light, turning off the light, adjusting brightness and dimming, and adjusting the light color tone, corresponding to user semantic basic elements, such as straightforward instructions like turning on and off the light, adjusting brightness and dimming, and adjusting the light color tone, or indirect semantic instructions like going to the toilet, going to the living room, or going to the kitchen. The second-type instructions are the dependent instructions set for each basic element in the second analysis result, such as ambient light status, electrical equipment status, sensor data analysis results, image analysis results, etc., which are related to the first-type instructions. The system establishes mutual or unidirectional dependencies between instructions. Based on the first and second analysis results, multiple heuristics are determined for each first-type instruction in each lighting functional unit. These heuristics include the dependency between first-type and second-type instructions, the relationship between first-type instructions and processing units, and the relationship between first-type instructions and lighting functional units, facilitating the selection of first-type instructions. The first-type instructions in each lighting functional unit are sorted multiple times. During each sort, one heuristic is selected as the sorting comparison value according to its priority, resulting in a sorted sequence of first-type instructions and generating lighting control instructions. This allows for the selection of first-type instructions with higher heuristic priorities. Thus, the system utilizes multiple data points from the first and second analysis results to jointly determine lighting control instructions, enabling the prediction and verification of user intentions, and determining the corresponding lighting control instructions to control the lighting equipment based on the prediction and verification results.

[0095] In a preferred embodiment, refer to Figure 3 The first type of instruction in each lighting functional unit performs multiple sequencing steps, including:

[0096] Step S1: Establish a first sorting sequence using all first-type instructions in each lighting functional unit, and select the heuristic with the highest priority as the first heuristic.

[0097] Step S2: Sort the first type of instructions a second time according to the value of the first heuristic, and divide the second sorted first type of instructions into multiple instruction subsequences according to the value of the heuristic, to obtain multiple sorted instruction subsequences;

[0098] Step S3: Take the optimal instruction subsequence corresponding to the optimal heuristic among the multiple sorted instruction subsequences as the first sorting sequence, and select the heuristic with the next priority as the first heuristic. Return to step S2 and continue until the sorting based on all heuristics is completed.

[0099] The working principle and beneficial effects of the above technical solution are as follows: When the first type of instructions in each lighting functional unit are executed multiple times, step S1 is used to establish a first sorting sequence using all the first type of instructions in each lighting functional unit, and the heuristic with the highest priority is selected as the first heuristic. Step S2 is used to sort the first type of instructions a second time according to the value of the first heuristic, and the first type of instructions after the second sorting are divided into multiple instruction subsequences according to the value of the heuristic, resulting in multiple sorted instruction subsequences. Step S3 is used as the first sorting sequence, and the heuristic with the next higher priority is selected as the first heuristic. Step S2 is then returned to be executed until the sorting based on all heuristics is completed. This achieves multiple sorting of the execution of the first type of instructions, resulting in a more scientific and accurate instruction sorting result.

[0100] In a preferred embodiment, the path prediction unit performs the following steps:

[0101] Based on the results of the first analysis, the keywords are determined, and based on the pre-defined mapping relationship between the keywords and each geographical location in the room, the target location corresponding to each keyword is determined.

[0102] The target location corresponding to each keyword is determined by statistical analysis, and the target location that appears most frequently is selected as the target location for movement.

[0103] Based on the preset room layout, determine the path connection between the user's current location and the target location, and use it as the prediction result of the user's movement path;

[0104] The prediction results are sent to the lighting control module as a control reference.

[0105] The working principle and beneficial effects of the above technical solution are as follows: The path prediction unit determines the keywords based on the first analysis result, and determines the target location corresponding to each keyword according to the preset mapping relationship between the keywords and each geographical location in the room. For example, binding the keyword "toilet" to the actual location of the toilet in the room allows the system to know the user's spatial intention when the user mentions the keyword "toilet". The system then statistically analyzes the target locations corresponding to each keyword and determines the most frequently occurring target location as the movement target location. For example, for the user saying "I want to go to the toilet to relieve myself", there are corresponding target locations: toilet - corresponding toilet location, relieve myself - corresponding toilet location, thus accurately judging the user's spatial intention. Based on the preset room layout, the system determines the path connection between the user's current location and the movement target location, and uses this as the prediction result of the user's movement path, realizing route planning prediction from the user's current location to the movement target location. The prediction result is sent as a control reference to the lighting control module, allowing the lighting control module to control the lighting equipment along the route according to the route planning prediction result. This ultimately achieves the scientific prediction of the user's movement route.

[0106] In a preferred embodiment, the model training unit includes:

[0107] The feature training subunit is used to acquire the voice data of the target user in daily communication and build the user's voice dataset. The voice dataset is then input into a general speech recognition model to train and extract the user's voice features.

[0108] The speech enhancement subunit is used to acquire the speech data of the target user reading a specific text, and to enhance the speech data by using the user's speech characteristics to obtain enhanced speech data.

[0109] The model training subunit is used to create a speech recognition model based on a preset keyword library and train it using enhanced speech data to obtain the trained speech recognition model.

[0110] The working principle and beneficial effects of the above technical solution are as follows: A feature training subunit acquires the target user's voice data from daily communication and establishes the user's voice dataset. This dataset is then input into a general speech recognition model for training and extraction of the user's voice features. A voice enhancement subunit acquires the target user's reading a specific text and enhances the reading data using the user's voice features to obtain enhanced voice data. A model training subunit creates a speech recognition model based on a pre-set keyword library and trains it using the enhanced voice data to obtain a trained speech recognition model. By extracting the user's voice features and enhancing the user's reading data to obtain enhanced voice data, the enhanced voice data highlights the user's voice characteristics in the reading of specific texts. Then, a speech recognition model based on the pre-set keyword library is created and trained using the enhanced voice data. The resulting speech recognition model exhibits good adaptability to user semantic recognition, achieving more accurate recognition results and reducing misrecognition.

[0111] In a preferred embodiment, the model training unit further includes:

[0112] The scenario creation subunit is used to establish the acoustic environment background for a specific speech recognition scenario and to obtain the background sound information under that acoustic environment background;

[0113] The speech synthesis subunit is used to synthesize synthesized speech data by using enhanced speech data and background sound information;

[0114] The context model training subunit is used to train a speech recognition model targeting a keyword library using synthesized speech data as the target input. After training, it establishes a mapping relationship between the model and the background sound environment to obtain a speech recognition model under that background sound environment.

[0115] The working principle and beneficial effects of the above technical solution are as follows: A scenario creation subunit establishes the acoustic environment background for a specific speech recognition scenario and acquires background sound information within that environment. A speech synthesis subunit synthesizes synthesized speech data using enhanced speech data and background sound information. A scenario model training subunit trains a speech recognition model targeting a keyword database using the synthesized speech data as the target input. After training, a mapping relationship is established between the model and the acoustic environment background to obtain the speech recognition model for that specific environment. This allows for the training of a speech recognition model for a user in a specific sound environment, resulting in a targeted speech recognition model that adapts well to specific environments. For example, speech recognition in a shower environment is prone to errors due to the background of shower noise, while a speech recognition model in a shower setting is more capable of performing speech recognition tasks.

[0116] In a preferred embodiment, the lighting control module includes:

[0117] The instruction receiving unit is used to receive lighting control instructions issued by the central control module;

[0118] The instruction analysis unit is used to analyze lighting control instructions, determine the equipment control channels of the lighting equipment that need to be controlled, and the control signals of the lighting equipment obtained from the analysis.

[0119] The instruction execution unit is used to control the relevant lighting equipment according to the equipment control channel and the control signals of the lighting equipment obtained by parsing.

[0120] The working principle and beneficial effects of the above technical solution are as follows: The instruction receiving unit receives lighting control instructions from the central control module; the instruction analysis unit analyzes the lighting control instructions to determine the device control channels of the lighting equipment to be controlled and the obtained control signals of the lighting equipment; the instruction execution unit controls the relevant lighting equipment according to the device control channels and the obtained control signals of the lighting equipment. This achieves unified control of all indoor lighting equipment based on the lighting control instructions issued by the central control module and has strong scalability.

[0121] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. An intelligent voice-controlled indoor lighting system, characterized in that, include: The sound recognition module is used to acquire sound information present indoors and analyze the sound information to obtain the first analysis result, including: The sound acquisition unit is used to acquire sound information present indoors; The voice extraction unit is used to extract voice data from the acquired sound information. The speech recognition unit is used to analyze human voice data to obtain semantic text through a preset speech recognition model; wherein, the speech recognition model is trained in advance using the user's speech training file; The keyword recognition unit is used to extract keywords from semantic text based on a preset keyword library and obtain the extraction results; The key sound extraction unit is used to identify key sounds in sound information based on a preset key sound frequency feature library to obtain the identification result; The first result generation unit is used to obtain the first analysis result based on the extraction result and the recognition result; The auxiliary recognition module is used to acquire indoor image data, electrical equipment data, and sensor data, and analyze them to obtain a second analysis result, including: The video acquisition unit is used to acquire video information appearing indoors; The video analysis unit is used to analyze video information, determine the human posture and gesture features, spatial relationship information of home appliances and lamps in the video, and generate video analysis results. The electrical analysis unit is used to acquire the operating status information of indoor electrical equipment and analyze the operating status to obtain electrical analysis results; The sensing analysis unit is used to acquire sensing data from indoor sensing devices and analyze the sensing data to obtain sensing data analysis results. The second result generation unit is used to generate a second analysis result based on the video analysis result, the electrical analysis result, and the sensor data analysis result. The central control module is used to determine the corresponding lighting control commands based on the first analysis result and the second analysis result; The lighting control module is used to control indoor lighting equipment according to lighting control commands.

2. The intelligent voice-controlled indoor lighting system according to claim 1, characterized in that, The central control module includes: The instruction selection unit is used to select the corresponding lighting control instruction from the first analysis result and the second analysis result, and to send the lighting control instruction to the lighting control module. The path prediction unit is used to determine the user's location indoors based on the first analysis result and the second analysis result, predict the user's movement path, obtain the prediction result, and send it to the lighting control module as a control reference. The model training unit is used to create a speech recognition model, and at the same time, it acquires the user's speech data and uses it as a speech training file to train the speech recognition model.

3. A smart voice-controlled indoor lighting system according to claim 2, characterized in that, The instruction selection unit execution includes: Obtain all first-type instructions and second-type instructions from the preset candidate instruction set of each lighting functional unit. The first-type instructions are the control instructions set for each basic element in the first analysis result, and the second-type instructions are the dependency instructions set for each basic element in the second analysis result. Based on the results of the first analysis and the second analysis, multiple heuristic quantities corresponding to each first type of instruction in each lighting functional unit are determined. The heuristic quantities include the dependency quantity between the first type of instruction and the second type of instruction, the relationship quantity between the first type of instruction and the processing unit, and the relationship quantity between the first type of instruction and the lighting functional unit. The first type of instructions in each lighting functional unit are sorted multiple times. During each sorting, one of the heuristic quantities is selected as the sorting comparison quantity according to the priority order of each heuristic quantity to obtain the sorted first type of instruction sequence and generate lighting control instructions.

4. A smart voice-controlled indoor lighting system according to claim 3, characterized in that, The first type of instructions in each lighting function unit performs multiple sequencing steps, including: Step S1: Establish a first sorting sequence using all first-type instructions in each lighting functional unit, and select the heuristic with the highest priority as the first heuristic. Step S2: Sort the first type of instructions a second time according to the value of the first heuristic, and divide the second sorted first type of instructions into multiple instruction subsequences according to the value of the heuristic, to obtain multiple sorted instruction subsequences; Step S3: Take the optimal instruction subsequence corresponding to the optimal heuristic among the multiple sorted instruction subsequences as the first sorting sequence, and select the heuristic with the next priority as the first heuristic. Return to step S2 and continue until the sorting based on all heuristics is completed.

5. A smart voice-controlled indoor lighting system according to claim 2, characterized in that, The path prediction unit performs the following steps: Based on the results of the first analysis, the keywords are determined, and based on the pre-defined mapping relationship between the keywords and each geographical location in the room, the target location corresponding to each keyword is determined. The target location corresponding to each keyword is determined by statistical analysis, and the target location that appears most frequently is selected as the target location for movement. Based on the preset room layout, determine the path connection between the user's current location and the target location, and use it as the prediction result of the user's movement path; The prediction results are sent to the lighting control module as a control reference.

6. A smart voice-controlled indoor lighting system according to claim 2, characterized in that, The model training unit includes: The feature training subunit is used to acquire the voice data of the target user in daily communication and build the user's voice dataset. The voice dataset is then input into a general speech recognition model to train and extract the user's voice features. The speech enhancement subunit is used to acquire the speech data of the target user reading a specific text, and to enhance the speech data by using the user's speech characteristics to obtain enhanced speech data. The model training subunit is used to create a speech recognition model based on a preset keyword library and train it using enhanced speech data to obtain the trained speech recognition model.

7. A smart voice-controlled indoor lighting system according to claim 6, characterized in that, The model training unit also includes: The scenario creation subunit is used to establish the acoustic environment background for a specific speech recognition scenario and to obtain the background sound information under that acoustic environment background; The speech synthesis subunit is used to synthesize synthesized speech data by using enhanced speech data and background sound information; The scenario model training subunit is used to train a speech recognition model targeting a keyword library using synthesized speech data as the target input. After training, it establishes a mapping relationship between the model and the background sound environment to obtain a speech recognition model under that background sound environment.

8. A smart voice-controlled indoor lighting system according to claim 1, characterized in that, The lighting control module includes: The instruction receiving unit is used to receive lighting control instructions issued by the central control module; The instruction analysis unit is used to analyze lighting control instructions, determine the equipment control channels of the lighting equipment that need to be controlled, and the control signals of the lighting equipment obtained from the analysis. The instruction execution unit is used to control the relevant lighting equipment according to the equipment control channel and the control signals of the lighting equipment obtained by parsing.

Images