Massage armchair massage manipulation generation method based on natural language audio
By applying a technology that combines a large language model with a medical knowledge base to massage chairs, the technology enables the automatic generation and execution of massage scripts from natural language audio. This solves the problem of the inability to dynamically adjust the massage path and intensity of massage chairs, thus improving the intelligence and safety of massage chairs.
Patent Information
- Application Number
- CN202511634304.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-10
- Publication Date
- 2026-02-06
AI Technical Summary
Existing massage chairs cannot dynamically adjust the massage path, intensity, frequency, and rhythm according to individual differences, and the voice interaction system lacks natural language understanding capabilities, resulting in a monotonous user experience and health risks.
By employing a large language model (LLM-RAG architecture) combined with a medical knowledge base, and through natural language audio data acquisition, preprocessing, text conversion, knowledge retrieval, and script generation, personalized massage techniques are automatically generated and executed, and massage tasks are updated in real time.
It achieves an intelligent, personalized, and safe massage chair experience, and can dynamically adjust massage techniques based on user feedback to avoid errors and health risks.
Smart Images
Figure CN121478956A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application relates to a massage chair massage method based on natural language audio, and belongs to the technical field of artificial intelligence voice interaction. BACKGROUND
[0002] The existing massage chairs generally adopt a preset program mode, and users can only select from a limited number of massage modes through physical buttons or mobile APPs. The massage path, intensity, frequency and rhythm are all fixed templates, and cannot be dynamically adjusted according to individual differences (such as muscle fatigue degree, pain area, body shape, taboo area). This fixed control method not only has a single use experience, but also easily causes discomfort or risk when there are specific health conditions (such as postoperative rehabilitation, osteoporosis or pregnancy taboo).
[0003] In recent years, some massage chairs have introduced voice control functions, but most systems are still based on a "keyword-instruction" mapping mechanism. For example, when the user says "shoulder massage" or "increase intensity", the system can only trigger the corresponding preset instruction. Such methods lack deep understanding of natural language semantics, cannot accurately identify complex expressions or multi-intention requests, and cannot perform multi-round interactions according to the context, thereby limiting the level of intelligence.
[0004] At the same time, the rapid development of large language model (LLM) technology provides stronger semantic understanding and knowledge reasoning capabilities for intelligent devices. By combining medical knowledge bases, human anatomy knowledge and user individual data, an intelligent closed loop of "natural language input-knowledge retrieval-technique generation-device execution" can be realized. However, there is currently no public technology that can fully apply the LLM-RAG architecture to the massage chair field to realize the entire process from voice interaction to automatic generation and execution of personalized massage scripts.
[0005] In addition, voice interaction in a closed space (such as a massage chair cabin) also faces many engineering challenges, including speaker echo interference, environmental noise suppression, insufficient user sound source positioning accuracy, and high voice recognition delay; at the execution level, how to map the generated natural language results to the underlying executable CAN / UART control instructions, and ensure safety and real-time performance, also lacks systematic solutions.
[0006] In summary, the existing technology still has a gap between natural language understanding, knowledge retrieval, voice signal processing and massage action execution, making it difficult to achieve truly intelligent, personalized and secure massage experiences. SUMMARY
[0007] To address the problems existing in the prior art, this invention proposes a method for generating massage techniques for massage chairs based on natural language audio.
[0008] The technical solution of the present invention is as follows: On one hand, the present invention provides a method for generating massage chair massage techniques based on natural language audio, including the following steps: Collect and preprocess the user's natural language audio data; Convert the preprocessed natural language audio data into natural language text data; Construct a knowledge retrieval database for massage medicine; Retrieve massage medicine knowledge from the massage medicine knowledge retrieval database based on natural language text data; A massage script generation model is constructed to generate massage chair execution scripts based on natural language text data and retrieved massage medicine knowledge. The massage chair executes a script to perform massage tasks. During the massage task, the user's natural language audio data is collected in real time, and a massage chair update script is generated according to the above steps. The massage chair then parses the update script and updates the massage task.
[0009] Preferably, the natural language audio data is preprocessed sequentially using an adaptive echo cancellation algorithm and a noise suppression algorithm.
[0010] Preferably, the preprocessed natural language audio data is converted into natural language text data using an automatic speech recognition algorithm.
[0011] Preferably, the massage medicine knowledge retrieval database includes indication data, contraindication data, acupoint distribution data, meridian pathway data, intensity range data, and clinical case data.
[0012] Preferably, massage medicine knowledge is retrieved from the massage medicine knowledge retrieval database using a retrieval enhancement generative model based on natural language text data.
[0013] Preferably, the massage script generation model is built based on a large language model, outputs a JSON script and adopts JSONSchema constraints. The JSON script fields include the technique type, target muscle group, intensity level, airbag beat, chair back angle, heat application temperature and duration.
[0014] Preferably, based on medical rules, it is determined whether there is a contraindicated combination of adjacent techniques in the current JSON script. If so, any technique in the current adjacent techniques is removed.
[0015] In another aspect, the present application also provides a massage chair massage method generation system based on natural language audio, comprising a data acquisition module, a language text conversion module, a knowledge base construction module, a knowledge base retrieval module, an execution script generation module, a script analysis execution module, and an incremental update module. The data acquisition module is configured to acquire natural language audio data of a user and perform preprocessing. The language text conversion module is configured to convert the preprocessed natural language audio data into natural language text data. The knowledge base construction module is configured to construct a massage medical knowledge retrieval library. The knowledge base retrieval module is configured to retrieve massage medical knowledge from the massage medical knowledge retrieval library according to the natural language text data. The execution script generation module is configured to construct a massage script generation model, and generate a massage chair execution script according to the natural language text data and the retrieved massage medical knowledge. The script analysis execution module is configured to enable the massage chair to analyze and execute the script to perform a massage task. The incremental update module is configured to acquire natural language audio data of a user in real time during a massage task, and generate a massage chair update script according to the above modules, and enable the massage chair to analyze and update the update script to update the massage task.
[0016] In another aspect, the present application also provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method of the present application.
[0017] In another aspect, the present application also provides a computer readable storage medium having a computer program stored thereon, wherein the program is executable on a processor to implement the method of the present application.
[0018] The present application has the following advantages: 1. The present application retrieves a massage medical knowledge base in real time through the RAG framework, and combines meridian direction, acupoint distribution, contraindication, and rehabilitation guideline information to constrain and filter the LLM output result, thereby avoiding "hallucination" error generation.
[0019] 2. The present application adopts a structured JSON Schema output format, so that the generated massage method script can be analyzed into underlying control instructions in real time, and the execution sequence can be dynamically interrupted or modified according to user feedback, thereby realizing "optimization while execution". BRIEF DESCRIPTION OF DRAWINGS
[0020] Figure 1 The present application is a method flowchart. DETAILED DESCRIPTION
[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0022] It should be understood that the step numbers used in the text are for ease of description only and are not intended to limit the order in which the steps are performed.
[0023] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.
[0024] The terms “comprising” and “including” indicate the presence of the described feature, whole, step, operation, element and / or component, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.
[0025] The term “and / or” refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes these combinations.
[0026] See Figure 1 In some embodiments, a method for generating massage chair massage techniques based on natural language audio is proposed, including the following steps: Collect and preprocess the user's natural language audio data; Convert the preprocessed natural language audio data into natural language text data; Construct a knowledge retrieval database for massage medicine; Retrieve massage medicine knowledge from the massage medicine knowledge retrieval database based on natural language text data; A massage script generation model is constructed to generate massage chair execution scripts based on natural language text data and retrieved massage medicine knowledge. The massage chair executes a script to perform massage tasks. During the massage task, the user's natural language audio data is collected in real time, and a massage chair update script is generated according to the above steps. The massage chair then parses the update script and updates the massage task.
[0027] In one specific embodiment, the massage chair parses and executes the script, which is mapped to the massage chair operating system task queue. The massage chair then executes massage tasks according to the task queue. During the massage task, the user's natural language audio data is collected in real time, and an incremental update script for the massage chair is generated according to the above steps. The massage chair parses and executes the script to obtain incremental data and updates the incremental data to the current task queue.
[0028] In one specific embodiment, a microphone array and a DOA estimation algorithm are used to collect the user's natural language audio data, which can effectively locate the user's position and avoid interference from external human voices. In some embodiments, natural language audio data are preprocessed sequentially using an adaptive echo cancellation algorithm and a noise suppression algorithm.
[0029] In some embodiments, preprocessed natural language audio data is converted into natural language text data using an automatic speech recognition algorithm.
[0030] In some embodiments, the massage medicine knowledge retrieval base includes indication data, contraindication data, acupoint distribution data, meridian pathway data, intensity range data, and clinical case data.
[0031] In some embodiments, massage medicine knowledge is retrieved from a massage medicine knowledge retrieval database using a retrieval-enhanced generative model based on natural language text data.
[0032] In one specific embodiment, the retrieval enhancement generation model employs a dual-tower dense vector model to perform similarity recall between user text and knowledge base entries, returning Top-K paragraphs.
[0033] In some embodiments, the massage script generation model is built based on a large language model, outputs a JSON script and adopts JSON Schema constraints. The JSON script fields include the type of massage technique, target muscle group, intensity level, airbag beat, chair back angle, heat application temperature and duration.
[0034] In one specific embodiment, the JSON script example is as follows:
[0035] The script parser maps the script to a timing control table for motors / valves / push rods / heat therapy and writes it to the real-time operating system queue; it supports interrupts and dynamic reprogramming. In one specific embodiment, the results and execution status generated by the TTS algorithm are converted into natural language and broadcast through a speaker, using an emotional tone to enhance the interactive experience.
[0036] In one specific embodiment, user history scripts, subjective ratings, and physiological parameters (heart rate variability HRV) are encrypted and stored in the cloud or locally; and personalized recommendation weights are updated periodically through federated fine-tuning.
[0037] In one specific embodiment, the large language model adopts a 7B to 13B parameter large language model fine-tuned in the medical field, and the Prompt template includes "user description + search terms + output format constraints".
[0038] In one specific embodiment, the massage medicine knowledge retrieval base and the massage script generation model are both set up on cloud / edge computing nodes, and natural language text data is uploaded to the cloud / edge computing nodes to execute subsequent steps.
[0039] In some embodiments, medical rules are used to determine whether adjacent techniques in the current JSON script are a contraindicated combination of techniques. If so, any technique in the current adjacent techniques is removed.
[0040] In a specific embodiment, an example of a neck and shoulder fatigue scenario is provided: The user said, "I had a one-hour video conference this afternoon, and my left shoulder and neck are very sore. I want to focus on massaging the left side, but avoid the Fengchi acupoint." DOA determines that the user is located on the massage chair, thus meeting the target user criteria. AEC and NS optimize audio, and ASR converts speech to text. RAG recalls "Upper Trapezius Muscle Kneading Method" and "Contraindications for Fengchi Acupoint"; LLM script generation: Knead the left trapezius muscle at level 4 for 60 seconds → Apply acupressure to the left levator scapulae at level 3 for 30 seconds → Perform balance relaxation on the corresponding area on the right side at level 2; Pressure at the Fengchi acupoint coordinates = 0; The parser sends out a UART frame, the mechanism moves to the left shoulder, and begins segmented execution; When a user requests "a little stronger," the system incrementally updates the script, increasing the strength of the corresponding segment by 1, which takes effect immediately.
[0041] In some embodiments, a massage chair massage technique generation system based on natural language audio is proposed, including a data acquisition module, a language text conversion module, a knowledge base construction module, a knowledge base retrieval module, an execution script generation module, a script parsing and execution module, and an incremental update module; The data acquisition module is used to collect the user's natural language audio data and perform preprocessing; The language-to-text conversion module is used to convert preprocessed natural language audio data into natural language text data. The knowledge base construction module is used to build a massage medicine knowledge retrieval database; The knowledge base retrieval module is used to retrieve massage medicine knowledge from the massage medicine knowledge retrieval database based on natural language text data; The execution script generation module is used to construct a massage script generation model and generate massage chair execution scripts based on natural language text data and retrieved massage medicine knowledge. The script parsing and execution module is used to parse and execute the script of the massage chair and map it to the task queue of the massage chair operating system. The massage chair then executes massage tasks according to the task queue. The incremental update module is used to collect the user's natural language audio data in real time during the massage task, and generate an incremental update script for the massage chair according to the above module. The massage chair parses and executes the script to obtain incremental data and updates the incremental data to the current task queue.
[0042] In some embodiments, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method as described in any embodiment of the present invention.
[0043] In some embodiments, a computer-readable storage medium is provided on which a computer program is stored, which, when executed by a processor, implements the method as described in any embodiment of the present invention.
[0044] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, A and B simultaneously, or B alone. A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of singular or plural items. For example, at least one of a, b, and c can represent: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, and c can be single or multiple.
[0045] Those skilled in the art will recognize that the units and algorithm steps described in the embodiments disclosed herein can be implemented using electronic hardware, computer software, or a combination of electronic hardware and software. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0046] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.
[0047] In the several embodiments provided in this application, any function, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0048] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.
Claims
1. A method for generating massage chair massage techniques based on natural language audio, characterized in that, Includes the following steps: Collect and preprocess the user's natural language audio data; Convert the preprocessed natural language audio data into natural language text data; Construct a knowledge retrieval database for massage medicine; Retrieve massage medicine knowledge from the massage medicine knowledge retrieval database based on natural language text data; A massage script generation model is constructed to generate massage chair execution scripts based on natural language text data and retrieved massage medicine knowledge. The massage chair executes a script to perform massage tasks. During the massage task, the user's natural language audio data is collected in real time, and a massage chair update script is generated according to the above steps. The massage chair then parses the update script and updates the massage task.
2. The method for generating massage chair massage techniques based on natural language audio according to claim 1, characterized in that, The natural language audio data is preprocessed sequentially using an adaptive echo cancellation algorithm and a noise suppression algorithm.
3. The method for generating massage chair massage techniques based on natural language audio according to claim 1, characterized in that, The preprocessed natural language audio data is converted into natural language text data using an automatic speech recognition algorithm.
4. The method for generating massage chair massage techniques based on natural language audio according to claim 1, characterized in that, The massage medicine knowledge retrieval database includes indication data, contraindication data, acupoint distribution data, meridian pathway data, intensity range data, and clinical case data.
5. The method for generating massage chair massage techniques based on natural language audio according to claim 1, characterized in that, Based on natural language text data, a retrieval-enhanced generative model is used to retrieve massage medicine knowledge from a massage medicine knowledge retrieval database.
6. The method for generating massage chair massage techniques based on natural language audio according to claim 1, characterized in that, The massage script generation model is built based on a large language model, outputs a JSON script and adopts JSON Schema constraints. The JSON script fields include the type of massage technique, target muscle group, intensity level, airbag beat, chair back angle, heat application temperature and duration.
7. The method for generating massage chair massage techniques based on natural language audio according to claim 6, characterized in that, Based on medical rules, determine whether adjacent techniques in the current JSON script are a contraindicated combination. If so, remove any of the adjacent techniques.
8. A massage chair massage technique generation system based on natural language audio, characterized in that, It includes a data acquisition module, a language-to-text conversion module, a knowledge base construction module, a knowledge base retrieval module, an execution script generation module, a script parsing and execution module, and an incremental update module; The data acquisition module is used to collect the user's natural language audio data and perform preprocessing; The language-to-text conversion module is used to convert preprocessed natural language audio data into natural language text data. The knowledge base construction module is used to build a massage medicine knowledge retrieval database; The knowledge base retrieval module is used to retrieve massage medicine knowledge from the massage medicine knowledge retrieval database based on natural language text data; The execution script generation module is used to construct a massage script generation model and generate massage chair execution scripts based on natural language text data and retrieved massage medicine knowledge. The script parsing and execution module is used by the massage chair to parse and execute scripts to perform massage tasks. The incremental update module is used to collect the user's natural language audio data in real time during the massage task, and generate a massage chair update script based on the above module. The massage chair then parses the update script to update the massage task.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 7.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 7.