A Traditional Chinese Medicine Acupuncture, Tuina, and Massage Device and Temperature Control System

Abstract

This invention discloses a traditional Chinese medicine acupuncture and massage device and its temperature control system, belonging to the field of traditional Chinese medicine massage technology. The invention includes a massage component and a temperature control system. The temperature control system includes a temperature acquisition module, a central control module, an execution adjustment module, and a feedback calibration module. The massage component achieves precise displacement between the moxibustion and massage mechanisms via a dual-axis moving rail, automatically adapting to the patient's back condition with the help of a camera. The massage mechanism's telescopic sleeve, combined with a spring, provides elastic massage. A pre-stored, customizable target temperature range is used, and adjustment commands are generated through differential analysis, resulting in significant muscle relaxation. The device automatically performs temperature comparison, command generation, and parameter correction, enhancing the synergistic therapeutic effect of moxibustion and massage, and reducing the operational and monitoring burden on medical staff. Initial temperature control is achieved through a suction fan and a rotating ring of a toggle mechanism. The execution adjustment module links the two, achieving bidirectional temperature control through speed adjustment and vent opening control.

Description

Technical Field

[0001] This invention relates to the field of traditional Chinese medicine massage technology, and in particular to a traditional Chinese medicine acupuncture massage device and temperature control system. Background Technology

[0002] Clinically, many patients with spinal diseases, such as scoliosis and thoracolumbar fasciitis, often experience a feeling of tension and tightness in the affected area. Taking the chest and back as an example, through palpation, doctors can find many stubborn muscle knots on the inner side of the scapula and both sides of the spine. Traditional Chinese medicine acupuncture and massage, as important diagnostic and treatment methods, usually use kneading or plucking techniques to focus on loosening the muscle knots. This is time-consuming and laborious. The strength, frequency, and temperature of manual massage depend on the experience and judgment of the operator, making it difficult to achieve standardized and precise control. Uneven strength, excessively high or low temperature can easily affect the treatment effect and even cause safety hazards such as skin burns. Summary of the Invention

[0003] The purpose of this invention is to provide a traditional Chinese medicine acupuncture, massage, and acupressure device and a temperature control system to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a traditional Chinese medicine acupuncture and massage device, comprising a massage component, a lifting frame and a lifting column, wherein the massage component is fixedly connected to the bottom of the lifting frame, one side of the lifting frame is disposed on the side of the lifting column, and a movable base is fixedly connected to the bottom of the lifting column. The massage component includes an outer cover, and the interior of the outer cover is provided with no less than 6 sets of dual-axis moving rails. Each set of dual-axis moving rails is provided with a moxibustion mechanism, and the bottom of the moxibustion mechanism is fixedly connected to a massage mechanism. A camera is installed at the bottom of the lifting frame, with the camera angled directly below the massage component.

[0005] Furthermore, the top of the moxibustion mechanism is provided with a connecting frame, and a first motor is provided above the connecting frame. The top of the first motor is fixedly connected to the dual-axis moving rail, and the output end of the first motor is fixedly connected to the top of the connecting frame. Purifiers are provided on both sides of the top of the connecting frame, and the purifiers are connected to the interior of the moxibustion mechanism through a suction fan.

[0006] Furthermore, the moxibustion mechanism is internally equipped with a retaining frame and a movable tray. The movable tray is positioned above the retaining frame. An insertion slot is provided on one side of the moxibustion mechanism, through which the movable tray is inserted into the interior of the moxibustion mechanism. An moxa cake is placed inside the movable tray. An annular groove is provided on the exterior of the moxibustion mechanism, and a rotating ring is placed inside the annular groove. Heat dissipation vents with matching positions and sizes are provided on the rotating ring and the annular groove. A toggle mechanism is provided on one side of the moxibustion mechanism, located outside the rotating ring.

[0007] Furthermore, a second motor is provided inside the actuating mechanism, and an actuating arm is connected to the output shaft of the second motor. A pull bar is provided on the outer side of the rotating ring, and the end of the actuating arm away from the second motor is slidably connected to the pull bar.

[0008] Furthermore, the massage mechanism includes a connecting plate, which is fixedly connected to the inside of the moxibustion mechanism. A connecting leg is fixedly connected to the center of the bottom surface of the connecting plate. A telescopic sleeve is provided at the end of each connecting leg. A spring is arranged around the outside of the telescopic sleeve. A massage head is fixedly connected to the bottom end of the telescopic sleeve.

[0009] Furthermore, a temperature control system, applied to the aforementioned traditional Chinese medicine acupuncture and massage device, includes: The temperature acquisition module is installed on the inner wall of the moxibustion device and inside the massage head of the massage device. It is configured to collect the core temperature of moxibustion inside the moxibustion device and the actual perceived temperature on the patient's skin surface in real time. The central control module is configured to be electrically connected to the temperature acquisition module, the execution adjustment module and the feedback calibration module respectively. It pre-stores the target temperature range corresponding to different massage programs, receives real-time temperature data transmitted by the temperature acquisition module, and generates temperature adjustment commands through comparative analysis. The execution adjustment module is linked with the suction fan and toggle mechanism of the moxibustion device. It is configured to respond to the adjustment command of the central control module and realize the bidirectional adjustment of the moxibustion temperature by adjusting the working intensity of the suction fan and the size of the heat dissipation vent of the rotating ring. The feedback calibration module, in conjunction with the camera and temperature acquisition module, is configured to dynamically calibrate the temperature regulation effect based on images of the patient's back skin condition and real-time temperature data, ensuring that the temperature is always within the target temperature range. The temperature acquisition module includes a first temperature sensor and a second temperature sensor. The first temperature sensor is embedded in the inner wall of the moxibustion mechanism and is located near the heat transfer port of the moving tray. It is used to collect the core heating temperature inside the moxibustion mechanism. The second temperature sensor is integrated into the bottom end face of the massage head and uses a contact temperature measuring element to collect the actual perceived temperature of the patient's skin surface.

[0010] Furthermore, the central control module includes a temperature analysis unit, an instruction generation unit, a scheme storage unit, and a temperature prediction unit; The solution storage unit pre-stores at least 3 sets of target temperature ranges corresponding to different massage scenarios. The target temperature ranges include the core temperature range of moxibustion and the temperature range of body sensation. The target temperature ranges under different scenarios can be customized and modified by external terminals. The temperature analysis unit receives standardized temperature data and calculates the first difference between the moxibustion core temperature and the corresponding target core temperature, and the second difference between the perceived temperature and the corresponding target perceived temperature. When the first difference or the second difference exceeds the preset allowable deviation, the temperature adjustment logic is triggered. The instruction generation unit generates adjustment instructions based on the difference magnitude and direction. The adjustment instructions include the speed adjustment parameters of the suction fan and the rotation angle parameters of the rotating ring of the toggle mechanism. The speed adjustment parameters correspond to the working intensity of the suction fan from level 0 to 5, and the rotation angle parameters correspond to the opening degree of the heat dissipation vent from 0 to 100%. The temperature prediction unit constructs a temperature change prediction model based on the historical temperature data from the temperature acquisition module and the adjustment records from the execution adjustment module, and calculates the predicted stable temperature after temperature adjustment through the temperature change prediction model. When the predicted stable temperature exceeds the target temperature range, the temperature prediction unit corrects the adjustment command parameters in advance to achieve predictive temperature adjustment and reduce temperature fluctuations.

[0011] Furthermore, the execution adjustment module includes a fan drive unit and a rotary drive unit; The fan drive unit is electrically connected to the suction fan, and controls the motor speed of the suction fan based on the speed adjustment parameters to adjust the heat flow extraction rate inside the moxibustion mechanism; The rotating ring drive unit is electrically connected to the second motor of the toggle mechanism, and controls the rotation angle of the second motor based on the rotation angle parameter to control the size of the opening.

[0012] Furthermore, methods for constructing temperature change prediction models include: Historical temperature data of multiple acupoint areas collected by the temperature acquisition module of a traditional Chinese medicine acupuncture and massage device are obtained to establish an acupoint thermal characteristic database. The acupoint thermal characteristic database includes the skin thickness, thermal conductivity coefficient, and blood vessel distribution density of each acupoint. Based on the acupoint thermal characteristic database, the acupoint thermal response weights are calculated, a weighted covariance matrix is ​​constructed and the eigenvalues ​​are solved. The eigenvectors with a cumulative contribution rate ≥95% are selected to generate a personalized spatial basis function set. Integrate historical temperature data from the temperature acquisition module, massage operation history records from the execution adjustment module, and acupoint thermal characteristic database; extract preset key time points based on timestamps, calculate dynamic temperature gradient, weighted temperature summation, pressure-temperature sensitivity, and mode adaptation features, and construct a massage operation-temperature coupling feature vector; The massage operation-temperature coupled feature vector is concatenated with a personalized spatial basis function set as input. An improved LSTM forget gate is introduced to introduce a massage frequency rhythm factor. A 3-layer LSTM network and a dropout layer are used to capture the temperature time-series dependency and output low-dimensional time-series features. The low-dimensional temporal features and the massage operation-temperature coupling feature vector are fused as input to construct a 4-layer ELM-AE stacked structure. The improved FLOO-CV algorithm is used to optimize the ridge parameters of each layer. The massage operation stability coefficient is introduced to correct the error calculation. The temperature of multiple acupoints is reconstructed based on the kernel function-personalized K-ELM. The reconstruction error is defined as the difference between the actual temperature and the reconstruction temperature. An operating state factor is introduced to decouple the error into the product of the spatial basis function and the time coefficient. The time coefficient is predicted by the ELM model, and the error compensation output is obtained by combining the personalized spatial basis function. A total loss function is constructed that includes temperature prediction error and massage rhythm penalty term. The parameters of the LSTM network are dynamically adjusted based on the variable parametric neurodynamic equation. The convergence speed of the variable parametric neurodynamic equation changes dynamically with the massage frequency. The parameters are iteratively updated by Euler discretization method. The temperature change prediction model is obtained by superimposing the ML-ELM reconstruction output with the error compensation output.

[0013] Furthermore, the feedback calibration module includes: The first calculation submodule is used to: determine the standard deviation of the back temperature distribution and the average back temperature based on real-time back temperature data; and determine the back heat distribution uniformity index based on the standard deviation of the back temperature distribution and the average back temperature. Analyze images of the skin condition on the back to determine the proportion of erythema, abnormal skin moisture, and cyanosis. Based on these proportions, determine the abnormal skin condition index. Calculate the real-time temperature difference ratio of the back based on real-time back temperature data; The skin thermal response coefficient is calculated based on the back heat distribution uniformity index, skin condition abnormality index, and back real-time temperature difference ratio. ; in, The skin thermal response coefficient; These are the weighting coefficients; To continuously adjust the time; The thermal relaxation time constant of the organization; The ratio of real-time temperature difference on the back; This is an index of abnormal skin condition; It is the heat distribution uniformity index; The second calculation submodule is used to calculate the adaptive temperature correction based on the skin thermal response coefficient. in, This is an adaptive temperature correction amount; It is the biological rhythm cycle; To control the cycle; This is the proportionality coefficient; The integral coefficient; These are the differential coefficients; To modulate the amplitude of physiological rhythms; For at a certain point in time The skin thermal response coefficient; The calibration submodule is used to determine the final temperature setpoint based on the real-time back temperature and the adaptive temperature correction amount; and to dynamically calibrate the temperature regulation effect based on the final temperature setpoint.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. The massage component of this invention achieves precise displacement between the moxibustion mechanism and the massage mechanism through a dual-axis moving rail. It automatically adapts to the patient's back condition with the help of a camera. The temperature control system is equipped with dual sensors to collect the core temperature of moxibustion and the skin's perceived temperature, breaking through the limitations of manual adjustment in traditional equipment. It achieves automatic and precise positioning of the massage and moxibustion position. Furthermore, it avoids single data deviation through dual-dimensional temperature measurement, providing comprehensive and reliable data support for temperature adjustment, effectively preventing burns or insufficient therapeutic effect, and greatly improving the safety, accuracy, and comfort of treatment. It is suitable for patients of different body types and various spinal disease treatment needs.

[0015] 2. This invention uses a first motor to drive the moxibustion mechanism and the massage mechanism to rotate. The telescopic sleeve of the massage mechanism, in conjunction with a spring, provides elastic massage. The central control module of the temperature control system integrates multiple unit functions, pre-stores a customizable target temperature range, generates adjustment commands through difference analysis, and achieves predictive adjustment by combining a temperature prediction model. This makes the massage movements more in line with the body's needs, significantly relieving muscle knots. At the same time, it does not require real-time manual intervention, automatically completing temperature comparison, command generation, and parameter correction. This solves the problems of low efficiency and strong lag in traditional manual temperature control, reduces temperature fluctuations, ensures stable heat output, enhances the synergistic therapeutic effect of moxibustion and massage, and reduces the operational and monitoring burden on medical staff.

[0016] 3. This invention achieves initial temperature control through the rotation of the suction fan and the actuation mechanism. The execution and adjustment module of the temperature control system links the two, achieving bidirectional temperature control through speed adjustment and heat dissipation vent opening control. The feedback calibration module dynamically calibrates the temperature based on the skin condition collected by the camera. The dual-path adjustment mechanism significantly improves the temperature control accuracy and response speed. The dynamic calibration function ensures that the temperature is adapted to individual patient differences, keeping the moxibustion temperature stable within the target range. At the same time, it does not require additional complex components, adapts to the original structure of the equipment, simplifies the overall design and reduces production costs. It is easy to operate and has strong universality, suitable for various TCM diagnosis and treatment scenarios. Attached Figure Description

[0017] Figure 1This is a schematic diagram of the overall structure of the traditional Chinese medicine acupuncture and massage device of the present invention; Figure 2 This is a schematic diagram of the massage component structure of the present invention; Figure 3 This is a schematic diagram of the external structure of the moxibustion mechanism of the present invention; Figure 4 This is a schematic diagram of the internal structure of the moxibustion mechanism of the present invention; Figure 5 This is a schematic diagram of the massage mechanism structure of the present invention; Figure 6 This is a schematic diagram of the actuation mechanism of the present invention; Figure 7 This is a schematic diagram of the temperature control process of the present invention.

[0018] In the diagram: 1. Massage component; 11. Outer cover; 12. Dual-axis moving rail; 13. Moxibustion mechanism; 131. First motor; 132. Connecting frame; 133. Purifier; 134. Fan; 135. Moving tray; 136. Clip-on frame; 137. Rotary ring; 138. Actuating mechanism; 1381. Second motor; 1382. Actuating arm; 1383. Pull bar; 139. Heat dissipation vent; 14. Massage mechanism; 141. Connecting plate; 142. Connecting support leg; 143. Telescopic sleeve; 144. Massage head; 145. Spring; 2. Lifting frame; 21. Camera; 3. Lifting column; 4. Moving base; 5. Moxa cake. Detailed Implementation

[0019] 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.

[0020] Please see Figure 1-7 The present invention provides the following technical solutions: A traditional Chinese medicine acupuncture and massage device includes a massage component 1, a lifting frame 2 and a lifting column 3. The massage component 1 is fixedly connected to the bottom of the lifting frame 2. One side of the lifting frame 2 is located on the side of the lifting column 3. The lifting column 3 is equipped with a drive component for driving the lifting frame 2 to move horizontally. A movable base 4 is fixedly connected to the bottom of the lifting column 3. The massage component 1 includes an outer cover 11. Inside the outer cover 11, there are at least 6 sets of dual-axis moving rails 12. Each set of dual-axis moving rails 12 is equipped with a moxibustion mechanism 13. The dual-axis moving rail 12 includes two parallel horizontal rails and one vertical rail. The top of the horizontal rail is fixedly connected to the inner top of the outer cover 11. The two ends of the vertical rail are slidably connected to the two horizontal rails through an electrically driven moving carriage assembly. The top of the moxibustion mechanism 13 is slidably connected to the vertical rail through an electrically driven moving carriage assembly. The bottom of the moxibustion mechanism 13 is fixedly connected to a massage mechanism 14. The positions of the moxibustion mechanism 13 and the massage mechanism 14 can be driven by the dual-axis moving rails 12, so that the position of the moxibustion and massage can be adjusted according to the patient's condition. A camera 21 is installed at the bottom of the lifting frame 2. The camera 21 is positioned directly below the massage component 1. The image of the patient's back captured by the camera 21 is combined with the massage plan selected by the medical staff. The positions of the moxibustion mechanism 13 and the massage mechanism 14 are automatically adjusted by the dual-axis moving rail 12 to automatically adapt to the patient's back condition and massage needs.

[0021] In the above embodiments, the massage components, lifting frame, and lifting column are modularly designed and combined with a mobile base to enable flexible movement of the device, meeting the needs of different treatment scenarios. The dual-axis moving rail adopts a combination structure of horizontal and vertical rails, combined with an electric-driven moving vehicle component, to achieve precise displacement of the moxibustion mechanism and the massage mechanism on the horizontal plane, breaking through the limitations of traditional manual adjustment. After the camera captures the image of the patient's back, the system can automatically match the massage plan and adjust the position of the mechanism without the need for medical staff to manually position it. This not only reduces the difficulty of operation, but also adapts to the back contours of patients of different body types, ensuring that the massage and moxibustion act on precise acupoints and improve the treatment effect. It is especially suitable for the treatment of precise areas such as both sides of the spine and the inner side of the scapula, and has greater universality.

[0022] A connecting frame 132 is provided on the top of the moxibustion mechanism 13. A first motor 131 is provided above the connecting frame 132. The top of the first motor 131 is fixedly connected to the dual-axis moving rail 12. The output end of the first motor 131 is fixedly connected to the top of the connecting frame 132. Purifiers 133 are provided on both sides of the top of the connecting frame 132. The purifiers 133 are connected to the interior of the moxibustion mechanism 13 through a suction fan 134. By driving the suction fan 134, some of the heat flow in the moxibustion mechanism 13 can be extracted. By adjusting the working intensity of the suction fan 134, the temperature of the moxibustion mechanism 13 can be controlled. The purifiers 133 can filter the smoke and dust in the extracted heat flow to a certain extent.

[0023] In the above embodiment, the first motor at the top of the moxibustion mechanism can drive the entire structure to rotate. Combined with the position adjustment of the dual-axis moving rail, it can realize multi-dimensional massage movements, resulting in a more significant effect of relieving muscle knots. The linkage design of the purifier and the suction fan on both sides of the connecting frame not only solves the problem of smoke and dust pollution during moxibustion, but also controls the amount of heat discharged by adjusting the working intensity of the suction fan, thus achieving preliminary adjustment of the moxibustion temperature. This integrated purification and temperature control design not only ensures the cleanliness of the treatment environment, but also avoids scalding patients due to excessively high temperatures or affecting the therapeutic effect due to excessively low temperatures, thus improving the safety and comfort of treatment. At the same time, it simplifies the equipment structure and reduces the investment in additional temperature control devices.

[0024] The moxibustion mechanism 13 is internally equipped with a retainer 136 and a movable tray 135. The movable tray 135 is positioned above the retainer 136. An insertion slot is provided on one side of the moxibustion mechanism 13, through which the movable tray 135 is inserted into the interior of the moxibustion mechanism 13. A moxa cake 5 is placed inside the movable tray 135. A pull handle is provided on the outside of the movable tray 135. A heat transfer port is provided at the bottom of the movable tray 135. An annular groove is provided on the outside of the moxibustion mechanism 13, inside which a rotating ring 137 is provided. Heat dissipation ports 139 with matching positions and sizes are provided on the rotating ring 137 and the annular groove. A toggle mechanism 138 is provided on one side of the moxibustion mechanism 13, outside of the rotating ring 137. When moxibustion is needed for the patient, the lit moxa cake 5 is placed in the movable tray 135 and then pushed into the interior of the moxibustion mechanism 13. This allows for simultaneous moxibustion when the massage component 1 descends to the patient's back for massage.

[0025] In the above embodiments, the movable tray adopts a pull-out design, which, together with the pull handle, allows for quick replacement of moxa cakes, making operation convenient. The rotating ring and annular groove on the outside of the moxibustion mechanism achieve secondary temperature regulation through the misalignment of the heat dissipation vents, complementing the temperature control function of the suction fan and further improving the temperature control accuracy. The locking bracket limits the movable tray, ensuring the stability of the moxa cake position during treatment. The design of the heat transfer port ensures the effective transfer of moxibustion heat, allowing the heat to be precisely applied to the patient's skin when moxibustion and massage are performed simultaneously, enhancing the synergistic therapeutic effect of moxibustion combined with massage, and is suitable for the treatment of various spinal diseases.

[0026] The actuating mechanism 138 is equipped with a second motor 1381. The output shaft of the second motor 1381 is connected to an actuating arm 1382. A pull bar 1383 is provided on the outer side of the rotating ring 137. The end of the actuating arm 1382 away from the second motor 1381 is slidably connected to the pull bar 1383. A sliding groove is provided inside the pull bar 1383. A sliding rod is provided on the end of the actuating arm 1382 away from the second motor 1381. The sliding rod is slidably connected inside the sliding groove. When the second motor 1381 drives the actuating arm 1382 to rotate, it can drive the pull bar 1383 to rotate, thereby driving the rotating ring 137 to rotate. This causes the rotating ring 137 and the annular groove to be misaligned, controlling the opening size of the heat dissipation vent 139 and controlling the moxibustion temperature inside the moxibustion mechanism 13.

[0027] In the above embodiment, the actuating mechanism drives the actuating arm through a second motor, and uses the sliding connection between the slide rod and the pull bar to drive the rotating ring to rotate, thereby realizing stepless adjustment of the size of the heat dissipation vent opening. The mechanical transmission structure is stable and reliable, not easily affected by high temperature environment, and has a longer service life. By controlling the opening degree of the heat dissipation vent, the heat exchange rate inside the moxibustion mechanism can be precisely adjusted. Combined with the heat flow extraction function of the suction fan, a two-way temperature control mechanism is formed to ensure that the moxibustion temperature remains stable during the treatment process, avoiding the impact of temperature fluctuation on the treatment effect. At the same time, it can be adjusted in real time according to the patient's tolerance, improving the personalization and safety of the treatment.

[0028] The massage mechanism 14 includes a connecting plate 141, which is fixedly connected to the inside of the moxibustion mechanism 13. The connecting plate 141 has a strip-shaped heat transfer port for heat transmission. A connecting leg 142 is fixedly connected to the center of the bottom surface of the connecting plate 141. Each end of the connecting leg 142 is provided with a telescopic sleeve 143. A spring 145 is arranged around the outside of the telescopic sleeve 143. A massage head 144 is fixedly connected to the bottom end of the telescopic sleeve 143. The telescopic sleeve 143 includes an inner cylinder and an outer cylinder that slide together. The top of the inner cylinder... The end is fixedly connected to the connecting leg 142, and the spring 145 is wrapped around the outside of the inner cylinder. The two ends of the spring 145 abut against the connecting leg 142 and the outer cylinder respectively. The massage head 144 is fixedly connected to the bottom of the outer cylinder. When the first motor 131 is working, it can drive the moxibustion mechanism 13 and the massage mechanism 14 to rotate. With the position adjustment of the dual-axis moving rail 12, it can massage the inner side of the patient's scapula and both sides of the spine, effectively loosening muscle knots. It is suitable for most spinal diseases. It is automatically controlled and easy to operate.

[0029] In the above embodiment, the telescopic sleeve of the massage mechanism adopts an inner and outer cylinder sliding sleeve structure, which, together with the surrounding spring, achieves elastic buffering, allowing the massage head to adapt to pressure when in contact with the patient's skin, avoiding discomfort caused by excessive force, while enhancing the comfort and fit of the massage. The strip-shaped heat transfer port on the connecting plate ensures that the heat from moxibustion can smoothly penetrate to the skin surface, achieving the synchronous effect of massage and moxibustion. The heat and mechanical massage work together to more easily loosen muscle knots and promote blood circulation. The first motor drives the moxibustion mechanism and the massage mechanism to rotate, and with the position adjustment of the dual-axis moving rail, it can achieve all-round massage of the sides of the spine, the inner side of the scapula, and other areas. There is no need for the patient to frequently adjust their body position, making it easy to operate and especially suitable for middle-aged and elderly patients and people with spinal diseases.

[0030] A temperature control system, applied in the aforementioned traditional Chinese medicine acupuncture and massage device, includes: The temperature acquisition module is located on the inner wall of the moxibustion mechanism 13 and inside the massage head 144 of the massage mechanism 14. It is configured to collect the core temperature of moxibustion inside the moxibustion mechanism 13 and the actual perceived temperature on the patient's skin surface in real time. The central control module is configured to be electrically connected to the temperature acquisition module, the execution adjustment module and the feedback calibration module respectively. It pre-stores the target temperature range corresponding to different massage programs, receives real-time temperature data transmitted by the temperature acquisition module, and generates temperature adjustment commands through comparative analysis. The execution adjustment module is linked with the suction fan 134 and the toggle mechanism 138 of the moxibustion mechanism 13. It is configured to respond to the adjustment command of the central control module and realize the bidirectional adjustment of the moxibustion temperature by adjusting the working intensity of the suction fan 134 and the opening size of the heat dissipation port 139 of the rotating ring 137. The feedback calibration module, linked with camera 21 and temperature acquisition module, is configured to dynamically calibrate the temperature regulation effect based on the patient's back skin condition image and real-time temperature data, ensuring that the temperature is always within the target temperature range. The temperature acquisition module includes a first temperature sensor and a second temperature sensor. The first temperature sensor is embedded in the inner wall of the moxibustion mechanism 13 and is located near the heat transfer port of the moving tray 135. It is used to collect the core heating temperature inside the moxibustion mechanism 13, and the sampling frequency is 10Hz. The second temperature sensor is integrated into the bottom end face of the massage head 144. It adopts a contact temperature measuring element and is used to collect the actual perceived temperature of the patient's skin surface. The sampling frequency is consistent with that of the first temperature sensor.

[0031] In the above embodiments, the temperature acquisition module adopts a dual-sensor design. The first temperature sensor accurately acquires the core temperature inside the moxibustion mechanism, and the second temperature sensor acquires the patient's skin temperature in real time. The dual-dimensional temperature measurement ensures the comprehensiveness of the data. The temperature acquisition is deeply integrated with the original structure of the device, eliminating the need for additional complex temperature measurement components. The structure is compact and does not affect the massage and moxibustion movements. By synchronously monitoring the core temperature and the perceived temperature, the problem of skin burns or insufficient therapeutic effects caused by adjusting only the internal temperature can be avoided. This makes the temperature control more in line with the patient's actual feeling and improves the safety and comfort of the treatment.

[0032] The central control module includes a temperature analysis unit, an instruction generation unit, a scheme storage unit, and a temperature prediction unit; The solution storage unit pre-stores at least 3 sets of target temperature ranges corresponding to different massage scenarios. The target temperature ranges include the core temperature range of moxibustion and the temperature range of body sensation. The target temperature ranges under different scenarios can be customized and modified through external terminals. The temperature analysis unit receives standardized temperature data and calculates the first difference between the core temperature of moxibustion and the corresponding target core temperature, and the second difference between the perceived temperature and the corresponding target perceived temperature. When the first difference or the second difference exceeds the preset allowable deviation, the temperature adjustment logic is triggered. The instruction generation unit generates adjustment instructions based on the difference magnitude and direction. The adjustment instructions include the speed adjustment parameters of the suction fan 134 and the rotation angle parameters of the rotating ring of the toggle mechanism 138. The speed adjustment parameters correspond to the working intensity of the suction fan 134 from 0 to 5 levels, and the rotation angle parameters correspond to the opening degree of the heat dissipation vent 139 from 0 to 100%. The temperature prediction unit constructs a temperature change prediction model based on historical temperature data from the temperature acquisition module and adjustment records from the execution adjustment module. The temperature change prediction model is then used to calculate the predicted stable temperature after temperature adjustment. When the predicted stable temperature exceeds the target temperature range, the temperature prediction unit corrects the adjustment command parameters in advance to achieve predictive temperature adjustment and reduce temperature fluctuations.

[0033] In the above embodiments, the scheme storage unit pre-stores multiple target temperature ranges, supports custom modification, and adapts to different massage scenarios and patient tolerance. The temperature analysis unit triggers adjustment logic by calculating temperature differences to ensure timely response when the temperature deviates. The instruction generation unit outputs precise rotation speed and angle parameters to achieve fine control of temperature control actions. The temperature prediction unit builds a model based on historical data and corrects the adjustment parameters in advance to achieve predictive temperature control, solving the lag problem of traditional feedback adjustment, reducing the temperature fluctuation range, and keeping the moxibustion temperature stable within the target range to ensure the consistency of treatment effects while reducing the monitoring pressure on medical staff.

[0034] The control module includes a fan drive unit and a rotary drive unit; The fan drive unit is electrically connected to the suction fan 134. Based on the speed adjustment parameter, the motor speed of the suction fan 134 is controlled to adjust the heat flow extraction rate inside the moxibustion mechanism 13. The speed and the heat flow extraction rate are positively correlated. The rotating ring drive unit is electrically connected to the second motor 1381 of the toggle mechanism 138. Based on the rotation angle parameter, the rotation angle of the second motor 1381 is controlled. The rotating ring 137 is driven to rotate through the toggle arm 1382, so that the rotating ring 137 is misaligned with the heat dissipation port 139 of the annular groove, thereby controlling the size of the opening. The rotation angle and the opening degree are linearly related.

[0035] In the above embodiments, the fan drive unit and the rotating ring drive unit of the execution adjustment module are linked to the suction fan and the second motor respectively to achieve the coordinated operation of the two temperature control methods. The fan speed is positively correlated with the heat flow extraction rate, and the rotating ring rotation angle is linearly correlated with the heat dissipation vent opening. The temperature control logic is clear, the adjustment accuracy is high, and the dual-path execution design can flexibly select the temperature control method according to the size of the temperature deviation. When the deviation is small, it can be finely adjusted by a single method, and when the deviation is large, it can be adjusted by dual-path coordination to improve the temperature control efficiency.

[0036] Methods for constructing temperature change prediction models include: Historical temperature data of multiple acupoint areas collected by the temperature acquisition module of a traditional Chinese medicine acupuncture and massage device are obtained to establish an acupoint thermal characteristic database. The acupoint thermal characteristic database includes the skin thickness, thermal conductivity coefficient, and blood vessel distribution density of each acupoint. Based on the acupoint thermal characteristic database, the acupoint thermal response weights are calculated, a weighted covariance matrix is ​​constructed and the eigenvalues ​​are solved. The eigenvectors with a cumulative contribution rate ≥95% are selected to generate a personalized spatial basis function set. Integrate historical temperature data from the temperature acquisition module, massage operation history records from the execution adjustment module, and acupoint thermal characteristic database; extract preset key time points based on timestamps, calculate dynamic temperature gradient, weighted temperature summation, pressure-temperature sensitivity, and mode adaptation features, and construct a massage operation-temperature coupling feature vector; The massage operation-temperature coupled feature vector is concatenated with a personalized spatial basis function set as input. An improved LSTM forget gate is introduced to introduce a massage frequency rhythm factor. A 3-layer LSTM network and a dropout layer are used to capture the temperature time-series dependency and output low-dimensional time-series features. The low-dimensional temporal features and the massage operation-temperature coupling feature vector are fused as input to construct a 4-layer ELM-AE stacked structure. The improved FLOO-CV algorithm is used to optimize the ridge parameters of each layer. The massage operation stability coefficient is introduced to correct the error calculation. The temperature of multiple acupoints is reconstructed based on the kernel function-personalized K-ELM. The reconstruction error is defined as the difference between the actual temperature and the reconstruction temperature. An operating state factor is introduced to decouple the error into the product of the spatial basis function and the time coefficient. The time coefficient is predicted by the ELM model, and the error compensation output is obtained by combining the personalized spatial basis function. A total loss function is constructed that includes temperature prediction error and massage rhythm penalty term. The parameters of the LSTM network are dynamically adjusted based on the variable parametric neurodynamic equation. The convergence speed of the variable parametric neurodynamic equation changes dynamically with the massage frequency. The parameters are iteratively updated by Euler discretization method. The temperature change prediction model is obtained by superimposing the ML-ELM reconstruction output with the error compensation output.

[0037] In this embodiment, 12 commonly used acupoints were selected, and the skin thickness, thermal conductivity, and blood vessel density of each acupoint area were measured experimentally and recorded as a physiological parameter matrix. ,in, For the first Physiological parameter vectors of acupoints ; For the first The skin thickness at each acupoint For the first The thermal conductivity coefficient of each acupoint Blood vessel density; physiological parameters were normalized.

[0038] In this embodiment, calculating the acupoint thermal response weights based on the acupoint thermal characteristic database includes: defining the acupoint thermal response weights. and normalize: ; These are the normalized physiological parameters; For the first Initial thermal response weights for each acupoint.

[0039] In this embodiment, the construction of the weighted covariance matrix includes: The construction process of the weighted covariance matrix deeply integrates the thermal characteristics and time series analysis of TCM acupoints: First, the system collects historical temperature data of 12 commonly used massage acupoints. This forms a spatiotemporal temperature matrix with 12 rows (acupoints) × L columns (time points). ,in Identify the spatial location of the i-th acupoint. ; This serves as a discrete-time index; subsequently, an L×L weighted covariance matrix is ​​constructed. Its elements The calculation was performed using a double-weighted approach—the outer layer summed the values ​​for 12 acupoints (each acupoint's contribution was weighted by a pre-calculated thermal response weight). Adjustment), inner layer for effective time window to The specific formula for performing temperature multiplication and accumulation is as follows: ,in, For time lag parameters, To ensure the computation window does not exceed its limits, this design cleverly achieves triple adaptability: First, the weights... (Based on normalized calculations of acupoint skin thickness, thermal conductivity, and blood vessel density) this makes densely vascularized areas (such as the Dazhui acupoint) contribute more to the covariance; secondly, dynamic time windows... Automatic adaptation to different lag scales ensures statistical reliability; finally, the dual summation structure separates the spatial characteristics and temporal correlation of acupoints, providing a precise time-frequency coupling basis for subsequent eigenvalue decomposition. The constructed... The matrix fully preserves the spatiotemporal evolution of the temperature field of multiple acupoints during traditional Chinese medicine massage, laying a mathematical foundation for extracting personalized spatial basis functions. This allows the temperature prediction model to both follow the meridian theory of traditional Chinese medicine and possess the rigor of modern signal processing.

[0040] In this embodiment, the eigenvalues ​​of the weighted covariance matrix are solved, and eigenvectors with a cumulative contribution rate ≥ 95% are selected to generate a personalized spatial basis function set, including: ; The characteristic decomposition equation; For eigenvalues; For feature vectors; select the first one The eigenvectors corresponding to the largest eigenvalues ​​(cumulative contribution rate ≥ 95%) are used to construct a personalized spatial basis function set: ;in, For the first The th eigenvector of the th feature vector One component; For the first Spatial basis functions at acupoints The value at; This is historical temperature data.

[0041] In this embodiment, the preset key time points are extracted based on timestamps, and the dynamic temperature gradient is calculated, including: ; For dynamic temperature gradient; The time interval for temperature sampling. For time windows; for Temperature values ​​measured by the time sensor; This represents the temperature sampling time interval; for Temperature value measured by the time sensor.

[0042] In this embodiment, the formula for calculating the weighted temperature accumulation includes: ,in, This is the weighted temperature cumulative value. This is the maximum massage frequency; for The frequency of massage at any given time; Exponentially decaying weights; For at a certain point in time Temperature value measured by a temperature sensor.

[0043] In this embodiment, the formula for calculating pressure-temperature sensitivity includes: ,in, This refers to the pressure-temperature sensitivity coefficient. , This refers to the change in massage pressure. This represents the temperature change between adjacent time steps; for Temperature values ​​measured by the time sensor; for Temperature value measured by the time sensor.

[0044] In this embodiment, the calculation formula for the pattern adaptation feature includes: ,in, Adapt feature vectors to the pattern; For massage mode identifiers (0: kneading mode, 1: pressing mode); For dynamic temperature gradient; This is the weighted cumulative temperature value.

[0045] In this embodiment, the dynamic temperature gradient, weighted temperature summation, pressure-temperature sensitivity, and pattern adaptation features are concatenated into a massage operation-temperature coupled feature vector, including: And supplement the current physiological parameters of acupoints. , The normalized physiological parameters of the current acupoint are: [Parameters to be filled in]; the coupled feature vector is: ;in, Basic feature vectors; For dynamic temperature gradient; This is the weighted temperature cumulative value; This refers to the pressure-temperature sensitivity coefficient. Adapt feature vectors to the pattern; These are the coupling feature vectors.

[0046] In this embodiment, the massage operation-temperature coupled feature vector is concatenated with a personalized spatial basis function set as input. An improved LSTM forgetting gate is introduced to incorporate a massage frequency rhythm factor. A three-layer LSTM network and dropout layers capture the temperature temporal dependency, outputting low-dimensional temporal features, including: The temperature prediction model employs a three-layer LSTM network architecture, with each layer containing 64 hidden units. Dropout layers with a dropout rate of 0.2 are set between layers to prevent overfitting, forming a hierarchical temporal feature extraction mechanism: the first LSTM layer receives the output of the embedding layer. ; The input vector for the LSTM; The number of basis functions in the space; From the coupled eigenvectors with space basis functions The first layer is assembled to capture short-term temperature change patterns and basic operations—temperature correlation. The second LSTM layer processes the abstract output of the first layer, identifies temperature dependencies at medium timescales, and integrates acupoint synergistic effects. The third LSTM layer further refines long-term temperature trends, generating low-dimensional temporal features. It fully characterizes the temperature evolution pattern; massage frequency rhythm factors are incorporated into the forgetting gates of all three-layer LSTMs, through formulas. To achieve; among which, Output for the forget gate; This is the current massage frequency; The baseline massage frequency; Here is the weight matrix of the LSTM forget gate; The hidden state vector from the previous time step; The input sequence is... This is the forget gate bias vector; The rhythm influence coefficient, The Sigmoid activation function enables the network to adaptively adjust its memory retention strategy based on the periodic changes in massage frequency. It strengthens short-term pattern memory when the operation rhythm is fast and enhances long-term trend tracking when the rhythm is stable, thereby accurately capturing the unique temperature-rhythm coupling relationship in the process of traditional Chinese massage and providing high-quality temporal features with traditional Chinese medicine physiotherapy characteristics for subsequent ML-ELM reconstruction.

[0047] In this embodiment, the low-dimensional temporal features are fused with the massage operation-temperature coupling feature vector as input to construct a 4-layer ELM-AE stacked structure. An improved FLOO-CV algorithm is used to optimize the ridge parameters of each layer, and a massage operation stability coefficient correction error calculation is introduced, including: Low-dimensional temporal features extracted by LSTM network Massage operation-temperature coupled feature vector containing dynamic temperature gradient, pressure sensitivity, and acupoint physiological characteristics. The data is then fused and used as input to ML-ELM. , The input vector is used for ML-ELM; a four-layer ELM-AE (Extreme Learning Machine-Autoencoder) stacked structure is constructed based on this. This structure adopts a layer-by-layer dimensionality reduction design, with the number of hidden layer nodes being 32, 16, 8, and 4 respectively, forming a hierarchical abstraction process from high-dimensional features to low-dimensional representations; to enhance the model's ability to perceive acupoint specificity, a modified sigmoid function is used as the activation function. ; To improve the sigmoid activation function; Weights are assigned to acupoint thermal responses to improve network sensitivity to densely vascularized or thermally sensitive areas. An improved FLOO-CV (fast leave-one-out cross-validation) algorithm is introduced to optimize ridge parameters at each layer, and the stability coefficient of massage operations is assessed. Dynamically assess the degree of fluctuation in massage frequency, among which, , The stability coefficient of the massage operation; for The frequency of massage at any given time; for The massage frequency at any given time; this coefficient is incorporated into the error calculation formula. This design allows the model to finely optimize parameters when the operation is stable and enhance robustness when the operation fluctuates, thereby adaptively balancing the accuracy and stability of temperature prediction, laying a solid foundation for subsequent multi-acupoint temperature reconstruction. , To improve FLOO-CV error; Let be the diagonal element of the hat matrix, and the lower bound of the denominator; The number of samples; This represents the stability coefficient of the massage operation.

[0048] In this embodiment, K-ELM based on kernel function personalization for multi-acupoint temperature reconstruction includes: The temperature prediction model uses K-ELM (Kernel Extreme Learning Machine) based on kernel function personalization to reconstruct the temperature field of multiple acupoints. The core is to design a kernel function that integrates the characteristics of acupoints: ;in, For kernel functions; For kernel width; For the first Thermal response weight of each acupoint; For feature vectors; introduce The normalized thermal response weights of the 12 acupoints were calculated. By incorporating a kernel function, similarity calculation can adaptively perceive the differences in thermal sensitivity among different acupoint regions. Through this kernel function, K-ELM maps the low-dimensional features processed by ELM-AE back to physical space, reconstructing 12 key acupoints in one go. At the present moment Temperature prediction matrix It fully characterizes the temperature field distribution on the back, providing a scientific basis for precise temperature control of different acupoint areas during TCM massage. It retains the nonlinear fitting ability of modern machine learning while fully incorporating the core concept of acupoint specificity in TCM theory.

[0049] In this embodiment, the reconstruction error is defined as the difference between the actual temperature and the reconstruction temperature, including: ; For the actual measured temperature; This represents the reconstruction error; To reconstruct the output temperature prediction values.

[0050] In this embodiment, an operational state factor is introduced to decouple the error into the product of a spatial basis function and a time coefficient, including: To accurately model the error characteristics of multi-acupoint temperature prediction, an operational state factor is introduced to achieve spatiotemporal decoupling of the error: First, the operational state factor is defined. ,in, This represents the current frictional pressure; Maximum permissible massage pressure; This is the current massage frequency; The baseline massage frequency; this factor comprehensively quantifies the influence of massage intensity and rhythm on tissue thermal response; based on this, the original reconstruction error is... Decoupling The product form of, where, Acupoints obtained through eigenvalue decomposition Personalized spatial basis functions characterize the inherent spatial pattern of thermal distribution in the acupoint region; This is a time-varying error coefficient, reflecting the dynamic evolution of the error during the massage process; while This term serves as an operational state modulation factor, enabling error compensation to adaptively sense operational intensity—when massage pressure or frequency increases. The system automatically enhances the error compensation amplitude to cope with more complex heat conduction changes under high-intensity operation. This decoupled design breaks through the limitations of traditional error processing, decomposes high-dimensional spatiotemporal errors into independently optimizable spatial patterns and temporal dynamics, and bridges the characteristics of traditional Chinese massage techniques with modern machine learning models through operation state factors. This enables the temperature prediction system to maintain high accuracy under different massage intensities, laying a theoretical foundation for the subsequent ELM model prediction of time coefficients and error compensation output. It significantly improves the adaptability and robustness of multi-acupoint temperature field reconstruction in traditional Chinese acupuncture and massage.

[0051] In this embodiment, the time coefficients are predicted using an ELM model, and the error compensation output is obtained by combining a personalized spatial basis function, including: The error compensation mechanism uses an Extreme Learning Machine (ELM) model to predict error time coefficients. Its core process is as follows: first, extract historical error time series. It captures short-term dynamic patterns of error evolution; and simultaneously integrates current operating parameters. ,in, Given the current frictional pressure, The current massage frequency is used to enable the prediction process to perceive the impact of operation intensity and rhythm on error. These five-dimensional feature vectors are input into a pre-trained ELM model, which quickly learns the complex relationship between historical errors and operation parameters through nonlinear mapping in a single hidden layer. The output layer generates the prediction error time coefficient for the current moment. This coefficient precisely quantifies the time-varying characteristics of the systematic error; ultimately, With personalized spatial basis functions and operating status factors Combined, an adaptive error compensation quantity is generated. This design enables dynamic correction of the baseline temperature prediction, allowing the system to proactively compensate for prediction errors. Especially when there are sudden changes in the intensity or frequency of massage operations, it significantly improves the accuracy and robustness of multi-acupoint temperature field reconstruction, ensuring that the treatment temperature remains stable within the target range.

[0052] In this embodiment, a total loss function is constructed that includes temperature prediction error and massage rhythm penalty term, including: ;in, This is the total loss function; For the actual measured temperature; This is a predicted temperature value; For error compensation output; This is the rhythm penalty coefficient. for The frequency of massage at any given time; For the first The frequency of massage at any given time.

[0053] In this embodiment, the parameters of the LSTM network are dynamically adjusted based on the variable parametric neural dynamics equation, including: Adaptive optimization of LSTM network parameters is achieved using variable-parameter neural dynamics equations. The core mechanism is: establishing dynamic differential equations. ,in Total loss function For model parameters gradient vector, To design convergence velocity parameters that dynamically change with the massage rhythm; ,in As a baseline convergence rate, This is the current massage frequency. The design uses a reference massage frequency; the convergence speed fluctuates periodically with the operating rhythm: when the massage frequency... Approximately reference frequency When it is an integer multiple of, Approaching 0, When the value is reduced to a baseline, the parameter updates become more refined and stable, making them suitable for capturing subtle changes in the temperature field; when Deviation When it is an integer multiple, Increase The parameters are increased to a maximum of 1.3 times the baseline value to accelerate parameter adjustment and adapt to sudden changes in operation. This dynamic mechanism enables the LSTM network to accurately optimize temperature prediction accuracy when the massage rhythm is stable and to converge quickly when the operation frequency changes drastically. It effectively balances the stability and adaptability of the model and ensures that the temperature control system can maintain prediction accuracy and real-time response capability under various massage techniques.

[0054] In this embodiment, parameter iterative updates are achieved using the Euler discretization method, including: To transform the continuous-time variable-parameter neural dynamics equations into a practically computable discrete iterative process, the Euler discretization method is used to achieve efficient updating of LSTM network parameters: based on time step... Constructing parameter iteration formulas ,in For model parameter vectors, The dynamic convergence speed parameters of the aforementioned design (adaptively changing with the massage rhythm) are used. loss function The gradient vector of the parameters, and This is a critical operational stability correction item. It quantifies the fluctuation of massage frequency within a local time window; the formula implements a triple adaptive mechanism: firstly, The base learning rate is dynamically adjusted based on the phase relationship between the massage frequency and the reference frequency; secondly, the operational stability coefficient... Through the exponential decay term Real-time adjustment of update amplitude – when operation is stable When the correction term is close to 1, it allows for sufficient parameter updates; when the operation fluctuates drastically... At that time, the correction term is significantly reduced, suppressing large jumps in parameters and preventing the model from being interfered with by noisy data; finally, the gradient term This design ensures that the update direction points to the path with the fastest loss reduction. This design enables the parameter optimization process to respond to the periodic characteristics of the massage rhythm and resist the interference caused by operational instability. In clinical trials, it significantly improves the convergence speed and generalization ability of the temperature prediction model under different massage techniques, providing a core guarantee for the robustness of the temperature control system.

[0055] In this embodiment, the temperature change prediction model is: ; This is the final temperature prediction value; To reconstruct the output temperature prediction value; This is the output for error compensation.

[0056] The working principle and beneficial effects of the above technical solution are as follows: By establishing a database of acupoint thermal characteristics and generating a personalized spatial basis function set, considering the characteristics of different acupoints such as skin thickness, thermal conductivity, and blood vessel distribution density, it can better adapt to individual differences and improve the accuracy of temperature prediction; by integrating historical temperature data, massage operation records, and the acupoint thermal characteristic database, a massage operation-temperature coupled feature vector is constructed, which comprehensively considers the relationship between massage operation and temperature changes, making the model more consistent with reality; by improving the LSTM forgetting gate and introducing the massage frequency rhythm factor, the temperature temporal dependency is captured through a 3-layer LSTM network and dropout layer, which can accurately grasp the temperature change pattern over time; a 4-layer ELM-AE stacked structure is adopted, combined with an improved FLOO-CV algorithm. The method optimizes ridge parameters, introduces a massage operation stability coefficient to correct error calculation, and uses a personalized K-ELM based on kernel functions to reconstruct the temperature of multiple acupoints, improving the accuracy of temperature reconstruction. An operation state factor is introduced to decouple the error, and error compensation output is obtained by combining the ELM model prediction time coefficient with a personalized spatial basis function, effectively reducing prediction error. A total objective function containing temperature prediction error and massage rhythm penalty terms is constructed, and the LSTM network parameters are dynamically adjusted based on the variable-parameter neurodynamic equation, enabling the model to dynamically optimize according to the massage frequency, enhancing the model's adaptability and robustness. The temperature change prediction model is obtained by superimposing the ML-ELM reconstruction output and the error compensation output. By integrating multiple optimization methods, accurate prediction of temperature changes in acupoint areas of traditional Chinese medicine acupuncture and massage devices is achieved.

[0057] The feedback calibration module includes: The first calculation submodule is used to: determine the standard deviation of the back temperature distribution and the average back temperature based on real-time back temperature data; and determine the back heat distribution uniformity index based on the standard deviation of the back temperature distribution and the average back temperature. Analyze images of the skin condition on the back to determine the proportion of erythema, abnormal skin moisture, and cyanosis. Based on these proportions, determine the abnormal skin condition index. Calculate the real-time temperature difference ratio of the back based on real-time back temperature data; The skin thermal response coefficient is calculated based on the back heat distribution uniformity index, skin condition abnormality index, and back real-time temperature difference ratio. ; in, The skin thermal response coefficient; These are the weighting coefficients; To continuously adjust the time; The thermal relaxation time constant of the organization; The ratio of real-time temperature difference on the back; This is an index of abnormal skin condition; It is the heat distribution uniformity index; The second calculation submodule is used to calculate the adaptive temperature correction based on the skin thermal response coefficient. in, This is an adaptive temperature correction amount; It is the biological rhythm cycle; To control the cycle; This is the proportionality coefficient; The integral coefficient; These are the differential coefficients; To modulate the amplitude of physiological rhythms; For at a certain point in time The skin thermal response coefficient; The calibration submodule is used to determine the final temperature setpoint based on the real-time back temperature and the adaptive temperature correction amount; and to dynamically calibrate the temperature regulation effect based on the final temperature setpoint.

[0058] In this embodiment, the back heat distribution uniformity index ,in To avoid dividing by zero for small constants; The standard deviation of the temperature distribution on the back; This represents the average temperature of the back.

[0059] In this embodiment, skin humidity abnormality refers to the absolute value of humidity deviation from the baseline value.

[0060] In this embodiment, the skin condition abnormality index is: ,in ; The percentage of the erythematous area on the back; Abnormal skin moisture level; This represents the percentage of areas with cyanosis.

[0061] In this embodiment, the real-time temperature difference ratio of the back is: ,in This represents the current average perceived temperature. For target perceived temperature, The ambient temperature.

[0062] In this embodiment, ,in As the reference gain, Variance; ,in For massage operation cycle ( ), The damping coefficient; This represents the variance of STRC over the most recent 100 time points.

[0063] In this embodiment, the final temperature setpoint The constraints are: ; in , Set the initial temperature value; To control the cycle.

[0064] The working principle and beneficial effects of the above technical solution are as follows: It comprehensively considers data from multiple dimensions, including back temperature distribution, skin condition, and real-time temperature difference ratio. By calculating the back heat distribution uniformity index, skin condition abnormality index, and back real-time temperature difference ratio, it comprehensively and accurately measures the skin's thermal response. It constructs a formula for calculating the skin thermal response coefficient, integrating multiple factors into a single indicator to achieve a quantitative assessment of the skin's thermal response, providing a scientific basis for subsequent temperature regulation. Based on the skin thermal response coefficient, it calculates an adaptive temperature correction amount. The formula considers proportional, integral, and derivative control, as well as biorhythm factors, enabling dynamic adjustment of the correction amount according to actual conditions, making temperature regulation more precise and flexible. Based on the adaptive temperature correction amount, it determines the final temperature setpoint, dynamically calibrating the temperature regulation effect, effectively improving the accuracy and stability of temperature control, and ensuring the system can adapt to different usage scenarios and individual differences.

[0065] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A traditional Chinese medicine acupuncture and massage device, comprising a massage assembly (1), a lifting frame (2), and a lifting column (3), characterized in that, The massage component (1) is fixedly connected to the bottom of the lifting frame (2), one side of the lifting frame (2) is located on the side of the lifting column (3), and the bottom of the lifting column (3) is fixedly connected to a movable base (4). The massage assembly (1) includes an outer cover (11), and the inner side of the outer cover (11) is provided with no less than 6 sets of dual-axis moving rails (12). Each set of dual-axis moving rails (12) is provided with a moxibustion mechanism (13), and the bottom of the moxibustion mechanism (13) is fixedly connected to a massage mechanism (14). A camera (21) is installed at the bottom of the lifting frame (2), and the camera (21) is positioned directly below the massage component (1).

2. The traditional Chinese medicine acupuncture and massage device as described in claim 1, characterized in that, The moxibustion mechanism (13) is provided with a connecting frame (132) on top, and a first motor (131) is provided above the connecting frame (132). The top of the first motor (131) is fixedly connected to the dual-axis moving rail (12), and the output end of the first motor (131) is fixedly connected to the top of the connecting frame (132). Purifiers (133) are provided on both sides of the top of the connecting frame (132). The purifiers (133) are connected to the interior of the moxibustion mechanism (13) through a suction fan (134).

3. The traditional Chinese medicine acupuncture and massage device as described in claim 2, characterized in that, The moxibustion mechanism (13) is provided with a card holder (136) and a movable tray (135) inside. The movable tray (135) is located above the card holder (136). An insertion slot is provided on one side of the moxibustion mechanism (13). The movable tray (135) is inserted into the interior of the moxibustion mechanism (13) through the insertion slot. A moxa cake (5) is provided inside the movable tray (135). An annular groove is provided on the outside of the moxibustion mechanism (13). A rotating ring (137) is provided inside the annular groove. Heat dissipation vents (139) with matching positions and sizes are provided on the rotating ring (137) and the annular groove. A toggle mechanism (138) is provided on one side of the moxibustion mechanism (13). The toggle mechanism (138) is located outside the rotating ring (137).

4. The traditional Chinese medicine acupuncture and massage device as described in claim 3, characterized in that, The actuating mechanism (138) is equipped with a second motor (1381) inside. A toggle arm (1382) is connected to the output shaft of the second motor (1381). A pull bar (1383) is provided on the outer side of the rotating ring (137). The end of the toggle arm (1382) away from the second motor (1381) is slidably connected to the pull bar (1383).

5. A traditional Chinese medicine acupuncture and massage device as described in claim 4, characterized in that, The massage mechanism (14) includes a connecting plate (141), which is fixedly connected to the inside of the moxibustion mechanism (13). A connecting leg (142) is fixedly connected to the middle of the bottom surface of the connecting plate (141). A telescopic sleeve (143) is provided at the end of each connecting leg (142). A spring (145) is arranged around the outside of the telescopic sleeve (143). A massage head (144) is fixedly connected to the bottom end of the telescopic sleeve (143).

6. A temperature control system, applied in a traditional Chinese medicine acupuncture and massage device as described in claim 1, characterized in that, include: The temperature acquisition module is set inside the inner wall of the moxibustion mechanism (13) and the massage head (144) of the massage mechanism (14), and is configured to collect the core temperature of moxibustion inside the moxibustion mechanism (13) and the actual perceived temperature of the patient's skin surface in real time. The central control module is configured to be electrically connected to the temperature acquisition module, the execution adjustment module and the feedback calibration module respectively. It pre-stores the target temperature range corresponding to different massage programs, receives real-time temperature data transmitted by the temperature acquisition module, and generates temperature adjustment commands through comparative analysis. The execution adjustment module is linked with the suction fan (134) and the toggle mechanism (138) of the moxibustion mechanism (13). It is configured to respond to the adjustment command of the central control module and realize the bidirectional adjustment of the moxibustion temperature by adjusting the working intensity of the suction fan (134) and the opening size of the heat dissipation port (139) of the rotating ring (137). The feedback calibration module, linked with the camera (21) and temperature acquisition module, is configured to dynamically calibrate the temperature regulation effect based on the patient's back skin condition image and real-time temperature data, ensuring that the temperature is always within the target temperature range. The temperature acquisition module includes a first temperature sensor and a second temperature sensor. The first temperature sensor is embedded in the inner wall of the moxibustion mechanism (13) and set near the heat transfer port of the moving tray (135) to collect the core heating temperature inside the moxibustion mechanism (13); the second temperature sensor is integrated into the bottom end face of the massage head (144) and adopts a contact temperature measuring element to collect the actual body temperature of the patient's skin surface.

7. A temperature control system as described in claim 6, characterized in that, The central control module includes a temperature analysis unit, an instruction generation unit, a scheme storage unit, and a temperature prediction unit. The solution storage unit pre-stores at least 3 sets of target temperature ranges corresponding to different massage scenarios. The target temperature ranges include the core temperature range of moxibustion and the temperature range of body sensation. The target temperature ranges under different scenarios can be customized and modified by external terminals. The temperature analysis unit receives standardized temperature data and calculates the first difference between the moxibustion core temperature and the corresponding target core temperature, and the second difference between the perceived temperature and the corresponding target perceived temperature. When the first difference or the second difference exceeds the preset allowable deviation, the temperature adjustment logic is triggered. The instruction generation unit generates adjustment instructions based on the difference magnitude and direction. The adjustment instructions include the speed adjustment parameters of the suction fan (134) and the rotation angle parameters of the rotating ring of the toggle mechanism (138). The speed adjustment parameters correspond to the working intensity of the suction fan (134) from 0 to 5, and the rotation angle parameters correspond to the opening degree of the heat dissipation port (139) from 0 to 100%. The temperature prediction unit constructs a temperature change prediction model based on the historical temperature data from the temperature acquisition module and the adjustment records from the execution adjustment module, and calculates the predicted stable temperature after temperature adjustment through the temperature change prediction model. When the predicted stable temperature exceeds the target temperature range, the temperature prediction unit corrects the adjustment command parameters in advance to achieve predictive temperature adjustment and reduce temperature fluctuations.

8. A temperature control system as described in claim 6, characterized in that, The execution adjustment module includes a fan drive unit and a rotary drive unit; The fan drive unit is electrically connected to the suction fan (134), and controls the motor speed of the suction fan (134) based on the speed adjustment parameter to realize the adjustment of the heat flow extraction rate inside the moxibustion mechanism (13); The rotating ring drive unit is electrically connected to the second motor (1381) of the toggle mechanism (138), and controls the rotation angle of the second motor (1381) based on the rotation angle parameter to control the size of the opening.

9. A temperature control system as described in claim 7, characterized in that, Methods for constructing temperature change prediction models include: Historical temperature data of multiple acupoint areas collected by the temperature acquisition module of a traditional Chinese medicine acupuncture and massage device are obtained to establish an acupoint thermal characteristic database. The acupoint thermal characteristic database includes the skin thickness, thermal conductivity coefficient, and blood vessel distribution density of each acupoint. Based on the acupoint thermal characteristic database, the acupoint thermal response weights are calculated, a weighted covariance matrix is ​​constructed and the eigenvalues ​​are solved. The eigenvectors with a cumulative contribution rate ≥95% are selected to generate a personalized spatial basis function set. Integrate historical temperature data from the temperature acquisition module, massage operation history records from the execution adjustment module, and acupoint thermal characteristic database; extract preset key time points based on timestamps, calculate dynamic temperature gradient, weighted temperature summation, pressure-temperature sensitivity, and mode adaptation features, and construct a massage operation-temperature coupling feature vector; The massage operation-temperature coupled feature vector is concatenated with a personalized spatial basis function set as input. An improved LSTM forget gate is introduced to introduce a massage frequency rhythm factor. A 3-layer LSTM network and a dropout layer are used to capture the temperature time-series dependency and output low-dimensional time-series features. The low-dimensional temporal features and the massage operation-temperature coupling feature vector are fused as input to construct a 4-layer ELM-AE stacked structure. The improved FLOO-CV algorithm is used to optimize the ridge parameters of each layer. The massage operation stability coefficient is introduced to correct the error calculation. The temperature of multiple acupoints is reconstructed based on the kernel function-personalized K-ELM. The reconstruction error is defined as the difference between the actual temperature and the reconstruction temperature. An operating state factor is introduced to decouple the error into the product of the spatial basis function and the time coefficient. The time coefficient is predicted by the ELM model, and the error compensation output is obtained by combining the personalized spatial basis function. A total loss function is constructed that includes temperature prediction error and massage rhythm penalty term. The parameters of the LSTM network are dynamically adjusted based on the variable parametric neurodynamic equation. The convergence speed of the variable parametric neurodynamic equation changes dynamically with the massage frequency. The parameters are iteratively updated by Euler discretization method. The temperature change prediction model is obtained by superimposing the ML-ELM reconstruction output with the error compensation output.

10. A temperature control system as described in claim 6, characterized in that, The feedback calibration module includes: The first calculation submodule is used to: determine the standard deviation of the back temperature distribution and the average back temperature based on real-time back temperature data; and determine the back heat distribution uniformity index based on the standard deviation of the back temperature distribution and the average back temperature. Analyze images of the skin condition on the back to determine the proportion of erythema, abnormal skin moisture, and cyanosis. Based on these proportions, determine the abnormal skin condition index. Calculate the real-time temperature difference ratio of the back based on real-time back temperature data; The skin thermal response coefficient is calculated based on the back heat distribution uniformity index, skin condition abnormality index, and back real-time temperature difference ratio. ； in, The skin thermal response coefficient; These are the weighting coefficients; To continuously adjust the time; The thermal relaxation time constant of the organization; The ratio of real-time temperature difference on the back; This is an index of abnormal skin condition; It is the heat distribution uniformity index; The second calculation submodule is used to calculate the adaptive temperature correction based on the skin thermal response coefficient. in, This is an adaptive temperature correction amount; It is the biological rhythm cycle; To control the cycle; This is the proportionality coefficient; The integral coefficient; These are the differential coefficients; To modulate the amplitude of physiological rhythms; For at a certain point in time The skin thermal response coefficient; The calibration submodule is used to determine the final temperature setpoint based on the real-time back temperature and the adaptive temperature correction amount; and to dynamically calibrate the temperature regulation effect based on the final temperature setpoint.

Images