Search

CN-121987409-A - Multi-mode fusion self-adaptive intelligent cold and hot touch Shu Zhuangzhi

CN121987409ACN 121987409 ACN121987409 ACN 121987409ACN-121987409-A

Abstract

The invention relates to the field of intelligent health, in particular to a multi-mode fusion self-adaptive intelligent cold and hot touch Shu Zhuangzhi. The temperature sensor comprises a touch Shu Mokuai, a control module and a multi-mode analysis module, wherein the touch Shu Mokuai adopts a TEC chip to realize refrigeration or heating, and can quickly and accurately regulate the temperature. The multi-mode analysis module is used for analyzing user feedback and physiological sensor data through intelligent algorithm fusion, generating emotion scores and personalized adjustment suggestions by combining a knowledge graph structure, and realizing closed-loop optimization of cold and hot touch experience. The control unit is responsible for system coordination, temperature control and safety monitoring. The device has the characteristics of being wearable, pollution-free, reusable and the like, is suitable for the fields of sports relaxation, emotion relief and personal health management, and provides scientific, safe and comfortable use experience.

Inventors

  • Dang Xuzhi
  • ZHANG ZHENMEI

Assignees

  • 山东第一医科大学附属省立医院(山东省立医院)

Dates

Publication Date
20260508
Application Date
20260112

Claims (10)

  1. 1. The multi-mode fusion self-adaptive intelligent cold and hot touch Shu Zhuangzhi is characterized by comprising a touch Shu Mokuai, a control module and a multi-mode analysis module; The touch Shu Mokuai is used for providing cold or heat for the target part; The control module comprises a PWM unit and a multi-mode interface unit, wherein the PWM unit is connected with the touch Shu Mokuai and used for controlling the working state of the touch module; The multi-mode analysis module is connected with the control module and is used for receiving the data of the control module for analysis, generating an adjustment strategy and returning the adjustment strategy to the control module; The multi-modal analysis module comprises an emotion analysis unit, a perception unit, an image analysis unit and a knowledge graph unit, wherein the emotion analysis unit constructs an inquiry type interaction scene, recognizes and evaluates subjective emotion states and preferences of a user in the use process in real time, quantifies the subjective emotion states and preferences into emotion scores, and transmits the emotion scores, problems in the interaction scene and user responses to the knowledge graph unit; the sensing unit collects multichannel physiological data, combines the application part and the device setting, realizes dynamic optimization of the sensor weight through the principal component analysis model, and transmits the sensor weight optimization result to the knowledge graph unit; the image analysis unit acquires image information of the target part, performs feature extraction and parameter quantization on the image information, generates an image analysis result and transmits the image analysis result to the knowledge graph unit; The knowledge graph unit builds a knowledge graph comprising various nodes and relations after pre-training, wherein the nodes comprise problem nodes, response nodes, sensor nodes, image nodes, emotion nodes, strategy nodes, execution result nodes, equipment mode nodes, treatment mode nodes and history treatment course nodes, the relations comprise co-occurrence relations, causal influences, parameter dependence and temperature control feedback, and the knowledge graph unit performs reasoning and searching similar states based on graph structures to generate an adjustment strategy and support incremental learning of a subsequent model.
  2. 2. The multi-modal fusion adaptive intelligent cold and hot touch Shu Zhuangzhi as set forth in claim 1, wherein the emotion analysis unit uses a BERT model, firstly, questions are presented and user responses are obtained, analog-to-digital conversion and word segmentation processing are performed on the user responses, the user responses are divided into { vocabulary, positions, sentence breaking } and then normalized, useless information is discarded, the remaining information is encoded, the BERT model is provided with a plurality of encoding layers, each layer receives vector codes of a word segmentation stage or a last encoding layer, emotion analysis is performed by referring to pre-training knowledge graph parameters, and emotion scores are generated through the processing of the plurality of encoding layers.
  3. 3. The multi-modal fusion adaptive intelligent cold and hot touch Shu Zhuangzhi as set forth in claim 2, wherein the BERT model combines the knowledge graph with the real-time sensor data to provide the user with an automatic tuning scheme based on feedback, and the user makes a final decision based on the automatic tuning scheme.
  4. 4. The multi-modal fusion adaptive intelligent cold and hot touch Shu Zhuangzhi as set forth in claim 1, wherein the image analysis unit processes and analyzes image information acquired by the image sensor using a CNN model, first acquires an original image of a user, then performs data enhancement and edge detection on the acquired image, extracts an ROI region and performs morphological filtering, and finally performs feature quantization on key information to form image acquisition information.
  5. 5. The multi-mode fusion self-adaptive intelligent cold and hot touch Shu Zhuangzhi according to claim 1, wherein the knowledge graph adopts a hierarchical association mechanism to infer, and the current strategy selection is output after weighted path search according to the sensor node, the image node, the emotion node and the strategy node.
  6. 6. The multi-modal fusion adaptive intelligent cold and hot touch Shu Zhuangzhi as set forth in claim 5, wherein after policy execution, the physiological variable, image change, temperature difference feedback, emotion fluctuation trend are written into the execution result node for updating policy success rate and realizing dynamic learning.
  7. 7. The multi-modal fusion adaptive intelligent cold and hot touch Shu Zhuangzhi as set forth in claim 1, wherein the knowledge-graph unit defines relationships as causal edges, temporal edges, weight edges, semantic edges, image feature edges.
  8. 8. The multi-mode integrated self-adaptive intelligent cold and hot touch Shu Zhuangzhi as set forth in claim 1, wherein the touch Shu Mokuai comprises a TEC cold plate and a radiator, the TEC cold plate comprises a contact surface and a radiating surface, the contact surface is refrigerated in a refrigeration mode, the radiating surface is heated, the radiator radiates heat, the contact surface is heated in a heating mode, the radiating surface is refrigerated, the radiator is not operated, and switching between the refrigeration mode and the heating mode is controlled by the direction of current flowing through the TEC cold plate.
  9. 9. The multi-mode integrated self-adaptive intelligent cold and hot touch Shu Zhuangzhi as set forth in claim 8, wherein the touch module further comprises a chip ADN8835 and a chip MAX6650, the chip ADN8835 controlling the operation of the TEC cold plate and the chip MAX6650 controlling the operation of the radiator.
  10. 10. The multi-mode fusion self-adaptive intelligent cold and hot touch Shu Zhuangzhi disclosed in claim 8, wherein a thermistor is arranged on the TEC cold sheet, the TEC cold sheet is connected with the PWM unit through the ADC unit, and the temperature of the TEC cold sheet is collected by the thermistor and fed back to the PWM unit.

Description

Multi-mode fusion self-adaptive intelligent cold and hot touch Shu Zhuangzhi Technical Field The invention relates to the field of intelligent health, in particular to a multi-mode fusion self-adaptive intelligent cold and hot touch Shu Zhuangzhi. Background The cold compress method commonly used for rehabilitation and exercise recovery at present mainly comprises a self-cooling chemical ice bag, a saline ice bag, an ethanol ice bag, a sleeve-belt type cold compress device and a cold therapy pressurizing device, and the main hot compress method mainly comprises hot water bag, hot wet compress, hot wet paste, microwave hot compress and the like. Therefore, the existing cold compress and hot compress modes are generally conducted by using disposable chemical materials or cold (heat) storage devices to achieve the purpose. The above method may cause pollution, and may require energy storage in advance, so there are problems of long time, poor efficiency, uneven effect, and pollution. In addition, the prior method is difficult to realize the variable and controllable cold (hot) compress effect, can not analyze the use effect and the user experience of the user, and proposes optimization suggestions. The semiconductor refrigerating plate is also called TEC chip, and can directly convert the input electric energy into temperature difference by utilizing the Peltier effect to realize refrigeration. The current is formed by the movement of electrons with different energy levels in the two conductors, when the electrons move from a high energy level to a low energy level, the surplus energy is released, and otherwise, the energy is absorbed. Energy is absorbed or released at the junction of two materials in the form of heat, so that the phenomenon that one end absorbs heat and the other end releases heat occurs at the two ends of the semiconductor refrigerating sheet. In addition, the cooling end and the heat release end are also interchanged when the direct current direction is changed. The semiconductor cold plate has small volume, high reliability and easy adjustment, and the semiconductor cold plate is used for refrigerating (heating), so that the effect is quick, no pollution is caused, and the semiconductor cold plate can be repeatedly used for many times. However, the existing medical or non-medical equipment based on semiconductor refrigeration is still mainly provided with 'fixed temperature output' and 'passive operation', and lacks intelligent capability. The main problems include incapability of collecting physiological index changes (such as heart rate, local temperature, blood flow, myoelectricity and the like) of a user in the using process, incapability of obtaining subjective experiences (such as comfort level, pain degree and cold and hot preference) of the user, and incapability of performing self-adaptive adjustment. Even if some devices have simple sensing input, the intelligent algorithm support such as pattern recognition, emotion analysis, individuation models and the like is lacking, and the history data accumulation and intelligent reasoning mechanism is lacking, so that the device cannot generate a differential strategy according to different crowds, different parts and different symptoms, and the accurate, safe and dynamically controllable cold and hot regulation effect is difficult to form. Particularly, the existing similar equipment generally lacks the 'multi-mode fusion' capability, namely, unified modeling analysis of language feedback, emotional state, image data (local swelling and temperature distribution) and physiological sensor data of a user cannot be realized, and a user model with continuous learning capability cannot be established. The lack of systematic analysis makes it difficult for the device to provide truly personalized treatments, nor can it form a feedback loop. Therefore, in the prior art, there is an urgent need for an intelligent cold and hot touch Shu Zhuangzhi with data acquisition capability, emotion analysis capability, multi-mode modeling capability and knowledge graph reasoning capability, so as to solve the problems of lack of intelligent adjustment, lack of self-learning capability, lack of user experience optimization mechanism and the like, thereby remarkably improving the safety, comfort, accuracy and effect consistency of cold and hot compress. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a multi-mode fusion self-adaptive intelligent cold and hot touch Shu Zhuangzhi to solve the problems that the existing touch and comfort device lacks intelligent adjustment, self-learning capability, user experience optimization mechanism and the like, thereby remarkably improving the safety, comfort, accuracy and effect consistency of cold and hot (hot) compress. In order to solve the technical problems, the invention adopts the technical scheme that the multi-mode fusion self-adaptive cold and hot touch device comprises a