A self-adaptive regulation method and system for a temperature controller based on acoustic signal feedback

Abstract

The application provides a self-adaptive regulation method and system of a warm instrument based on acoustic signal feedback, and relates to the technical field of a warm instrument, including the following steps: collecting mixed acoustic signals during the operation of the warm instrument through an acoustic sensor integrated in the internal part of a physiotherapy end of the warm instrument; the mixed acoustic signals include the vibration noise of the device itself and the interaction signal with the human tissue; processing the mixed acoustic signals to separate and extract the characteristic acoustic fingerprint; based on the characteristic acoustic fingerprint, inferring the current coupling state of the physiotherapy end and the human tissue; in response to the tissue coupling state, generating a first adjustment instruction for at least one output working parameter of the warm instrument to optimize the energy transmission efficiency between the physiotherapy end and the human tissue. The scheme can realize direct and accurate perception of the coupling state, improve the efficiency and stability of energy transmission, and enhance the safety and consistency of treatment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of warm-up instruments, in particular to a warm-up instrument adaptive regulation method and system based on acoustic signal feedback. BACKGROUND

[0002] The existing warm-up instrument (such as a physical therapy device based on ultrasonic wave, vibration, etc.) highly depends on the coupling state between the treatment head (physical therapy end) and the human skin / tissue. The ideal coupling can ensure that the energy is efficiently and uniformly transmitted to the target tissue layer; and the poor coupling (such as caused by the treatment head tilt, uneven contact pressure, skin wrinkle or lack of coupling agent / unevenness) will cause a large amount of energy reflection at the interface, which not only greatly reduces the treatment efficiency, but also may cause safety risks such as device damage or user scald due to energy accumulation in the treatment head or surface tissue.

[0003] At present, the monitoring of the coupling state mainly depends on two ways:

[0004] Operator experience dependent: the therapist subjectively judges by hand feeling, visual observation or inquiry of patient's feeling, which is extremely inaccurate and cannot realize real-time, quantitative feedback and regulation.

[0005] Simple electrical parameter feedback type: some devices indirectly infer the load change by monitoring the load current or output end impedance of the driving motor and other electrical parameters. However, this method has low sensitivity and specificity, and it is difficult to distinguish the real tissue coupling change from the fluctuation of the device itself running state, and it is also difficult to identify the complex coupling scene (such as different tissue types, different degrees of poor coupling)

[0006] In summary, how to realize real-time, accurate and non-invasive perception of the physical coupling state between the physical therapy end of the warm-up instrument and the human tissue, and automatically adjust the device output based on the state, to overcome the energy transmission efficiency fluctuation problem caused by the change of contact condition, needs to be solved urgently. SUMMARY

[0007] In view of the above defects or deficiencies in the prior art, the present application aims to provide a warm-up instrument adaptive regulation method and system based on acoustic signal feedback.

[0008] In a first aspect, the present application provides a warm-up instrument adaptive regulation method based on acoustic signal feedback, comprising the following steps:

[0009] The acoustic sensor integrated in the internal physical therapy end of the warm-up instrument collects the mixed acoustic signal during the operation of the warm-up instrument; the mixed acoustic signal includes the vibration noise of the device itself and the action signal interacting with the human tissue;

[0010] processing the mixed acoustic signal to separate and extract a characteristic acoustic fingerprint;

[0011] inferring a current tissue coupling state of the physiotherapy end and the human body based on the characteristic acoustic fingerprint;

[0012] in response to the tissue coupling state, generating a first adjustment instruction for at least one output working parameter of the thermophore to optimize the energy transmission efficiency between the physiotherapy end and the human tissue.

[0013] According to the technical scheme provided in the present application, the processing of the mixed acoustic signal to separate and extract a characteristic acoustic fingerprint comprises the following steps:

[0014] under the predetermined condition that the physiotherapy end of the thermophore is in an empty load and works, collecting and establishing a baseline acoustic fingerprint, which represents the vibration noise;

[0015] in a normal treatment process, collecting real-time mixed acoustic signals;

[0016] comparing the real-time mixed acoustic signal with the baseline acoustic fingerprint, and separating the characteristic acoustic fingerprint generated by the interaction between the physiotherapy end and the human tissue by calculating the difference between the two in the predetermined frequency domain.

[0017] According to the technical scheme provided in the present application, the inference of the current tissue coupling state of the physiotherapy end and the human tissue based on the characteristic acoustic fingerprint comprises the following steps:

[0018] matching the characteristic acoustic fingerprint with a preset acoustic coupling template database to obtain a target physical coupling scene corresponding to the characteristic acoustic fingerprint;

[0019] wherein the preset acoustic coupling template database includes a plurality of preset acoustic coupling templates and a physical coupling scene corresponding to each of the preset acoustic coupling templates, and the physical coupling scene includes ideal coupling, poor coupling, and crossing different tissue types.

[0020] the target physical coupling scene is used as the inferred tissue coupling state.

[0021] According to the technical scheme provided in the present application, the preset acoustic coupling template database is a multi-dimensional dynamic database; wherein each of the stored preset acoustic coupling templates is associated with at least one of the following dimensional labels:

[0022] user identification dimension, used to distinguish individual tissue characteristics of different users;

[0023] body part dimension, used to distinguish different anatomical sites of the physiotherapy end;

[0024] The treatment time sequence dimension is used to characterize the different treatment time stages from the start to the end of a single treatment session.

[0025] The step of matching the characteristic acoustic fingerprint with a preset acoustic coupling template database to obtain the target physical coupling scene corresponding to the characteristic acoustic fingerprint includes the following steps:

[0026] Based on the current user identifier, and / or the identified body part, and / or the current treatment time stage, a target template subset is selected from the preset acoustic coupling template database;

[0027] The real-time acquired acoustic fingerprint features are matched with preset acoustic coupling templates in the target template subset to determine the target physical coupling scenario.

[0028] According to the technical solution provided in this application, after matching the real-time acquired acoustic fingerprint with the preset acoustic coupling template in the target template subset, the following steps are further included:

[0029] Calculate the confidence score between the characteristic acoustic fingerprint and the best matching template in the target template subset;

[0030] Determining the target physical coupling scenario includes the following steps:

[0031] If the confidence score is higher than or equal to the preset confidence threshold, then the physical coupling scenario corresponding to the best matching template is determined as the target physical coupling scenario.

[0032] According to the technical solution provided in this application, after calculating the confidence score of the feature acoustic fingerprint and the best matching template in the target template subset, the method further includes the following steps:

[0033] If the confidence score is lower than the preset confidence threshold, the characteristic acoustic fingerprint is determined to be candidate abnormal data;

[0034] A continuous observation window is started for the candidate abnormal data, and the frequency and stability of similar abnormal data with similar characteristics are monitored within the observation window;

[0035] Based on the monitoring results, perform one of the following operations:

[0036] If, within the observation window, abnormal data with similar characteristics appear frequently and stably, it is determined that they constitute a new effective physical coupling scenario, and a new acoustic coupling template is generated based on the similar abnormal data and stored in the preset acoustic coupling template database.

[0037] If the abnormal data appears only at low frequency or transiently within the observation window, it is determined to be interference noise, and a preset safety control strategy is triggered.

[0038] According to the technical solution provided in this application, the following steps are also included:

[0039] Based on the characteristic acoustic fingerprint, the tissue temperature of the treatment area is calculated in real time.

[0040] The tissue temperature is compared with a preset target temperature range;

[0041] In response to the tissue temperature deviating from the preset target temperature range, a second adjustment command is generated for the operating parameters of the temperature transmitter to dynamically adjust the energy output and bring the tissue temperature back to the preset target temperature range.

[0042] Specifically, the tissue temperature is calculated by analyzing the mapping relationship between the frequency domain characteristics of the characteristic acoustic fingerprint and temperature-related acoustic parameters; the temperature-related acoustic parameters include sound velocity variation, sound attenuation coefficient, or resonant frequency shift.

[0043] According to the technical solution provided in this application, the following steps are also included:

[0044] Based on the tissue coupling state and the tissue temperature, a corresponding priority decision rule is obtained;

[0045] Based on the aforementioned priority decision rules, target adjustment instructions are generated to synergistically optimize energy transfer efficiency and thermal management objectives.

[0046] The priority decision rules include:

[0047] When the absolute value of the difference between the tissue temperature and the preset safety threshold is less than the first preset threshold, the second adjustment command is selected as the target adjustment command;

[0048] When the absolute value of the difference between the tissue temperature and the preset safety threshold is greater than or equal to the first preset threshold, the first adjustment instruction is selected as the target adjustment instruction.

[0049] According to the technical solution provided in this application, after processing the mixed acoustic signal, the following steps are also included:

[0050] From the characteristic acoustic fingerprint, acoustic elastic parameters for characterizing the physiological state of the tissue are extracted; the acoustic elastic parameters are obtained by analyzing the energy distribution changes or resonant peak shifts in specific frequency bands of the characteristic acoustic fingerprint.

[0051] Based on the changing trend of the acoustic elasticity parameters over the treatment time series, the physiological response intensity of human tissue during the treatment process can be obtained in real time.

[0052] After generating a first adjustment command for at least one output operating parameter of the temperature transmitter in response to the tissue coupling state, the method further includes the following steps:

[0053] Based on the intensity of the physiological response, the first adjustment instruction is modified to generate an optimized first adjustment instruction;

[0054] The correction rule is as follows:

[0055] When the physiological response intensity is lower than the expected response threshold, the adjustment range of the energy output parameter in the first adjustment command is increased;

[0056] When the physiological response intensity is higher than the expected response threshold, the adjustment range of the energy output parameter in the first adjustment command is reduced or maintained.

[0057] Secondly, this application proposes an adaptive control system for a temperature flow meter based on acoustic signal feedback, used to implement the adaptive control method for a temperature flow meter based on acoustic signal feedback as described above, including:

[0058] The acquisition module is configured to acquire mixed acoustic signals during the operation of the thermal oscillator via an acoustic sensor integrated inside the physiotherapy end of the thermal oscillator; the mixed acoustic signals include the vibration noise of the device itself and the interaction signals with human tissue.

[0059] An extraction module is configured to process the mixed acoustic signal to separate and extract characteristic acoustic fingerprints;

[0060] A judgment module is configured to infer the current coupling state between the physiotherapy terminal and human tissue based on the characteristic acoustic fingerprint.

[0061] A control module configured to generate a first adjustment command for at least one output operating parameter of the temperature transducer in response to the tissue coupling state, so as to optimize the energy transfer efficiency between the therapy end and the human tissue.

[0062] Compared with the prior art, the beneficial effects of this application are as follows:

[0063] I. Direct and accurate perception of coupling state: By analyzing the acoustic signals (rather than indirect electrical parameters) generated after the interaction between the physiotherapy end and human tissue, the characteristic acoustic fingerprint reflecting the physical properties of the interface can be directly obtained, thereby achieving high-precision, non-invasive real-time diagnosis of coupling state (such as ideal coupling, poor coupling, acting on different tissues, etc.).

[0064] Second, it improves the efficiency and stability of energy transfer: By generating the first adjustment command to optimize operating parameters (such as power, frequency, and amplitude) in real time, the system can actively compensate for energy loss caused by changes in coupling state. For example, it automatically increases the output to maintain an effective dose when slight coupling defects are detected, or reduces the output accordingly after the coupling is improved to avoid overdose, thereby ensuring that the efficiency of energy transfer to tissues remains at a high and stable level throughout the treatment process.

[0065] Third, it enhances the safety and consistency of treatment: By monitoring and responding immediately to adverse coupling conditions in real time, it effectively prevents the risk of local overheating caused by energy accumulation at the interface, protecting patients from potential burns and protecting the equipment from damage caused by reflected energy. Furthermore, this automated control reduces reliance on the operator's personal experience, making treatment effects more repeatable and consistent across different operators and treatment sessions.

[0066] Fourth, it lays the foundation for intelligent and personalized treatment: This solution transforms the coupling state from an uncontrollable variable into a monitorable, feedback-able, and controllable system parameter, providing core data input and control interfaces for subsequent more complex intelligent control (such as combining tissue type, individual user differences, etc.). Attached Figure Description

[0067] Figure 1 A schematic diagram illustrating the steps of the adaptive control method for a temperature and ventilation instrument based on acoustic signal feedback provided in an embodiment of this application. Detailed Implementation

[0068] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0069] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0070] Example 1

[0071] As mentioned in the background section, in view of the problems in the prior art, this application proposes an adaptive control method for a temperature and flux meter based on acoustic signal feedback, such as... Figure 1 As shown, it includes the following steps:

[0072] The acoustic sensors integrated into the physiotherapy end of the thermometer collect mixed acoustic signals during the operation of the thermometer; the mixed acoustic signals include the vibration noise of the device itself and the interaction signals with human tissue.

[0073] In this embodiment, the thermotherapy device specifically refers to a device that uses mechanical vibration energy (such as ultrasound, focused sound waves, or broadband vibration) to provide thermotherapy to human tissue. The therapy end is the treatment head that directly contacts the human body. Its core is a transducer (such as a piezoelectric ceramic transducer) used to convert electrical energy into mechanical vibration. To implement this method, one or more acoustic sensors are integrated inside the therapy end, adjacent to the transducer. These sensors are preferably piezoelectric accelerometers or microphones, and their function is to sense the vibration sound of the therapy end during operation. When the thermotherapy device is working, the transducer vibrates, and this vibration is transmitted to the human tissue through the therapy end. At this time, the acoustic sensor collects a mixed acoustic signal, which contains at least two parts: one is the vibration noise generated by the vibration of the device's own motor, transducer, and structure; the other is the feedback signal after the vibration energy interacts with human skin, muscles, and other tissues (such as reflection, transmission, absorption, and scattering). This signal is usually recorded by a data acquisition card in the form of a voltage-time series.

[0074] The mixed acoustic signals are processed to separate and extract characteristic acoustic fingerprints;

[0075] Specifically, the raw mixed signal cannot be used directly and must be processed. This step is completed in the device's embedded processor or a connected mobile terminal app. The primary goal of the processing is to extract stable and repeatable features that characterize the current "device-organization" interaction state, i.e., the "characteristic acoustic fingerprint." This fingerprint is not a single numerical value, but a multi-dimensional feature vector. Specific processing steps may include: signal preprocessing (such as bandpass filtering to remove irrelevant high-frequency noise and power frequency interference) and time-frequency analysis (such as short-time Fourier transform - STFT, converting a one-dimensional time signal into a two-dimensional time-spectrum graph). From the time-spectrum graph, a series of features such as resonant frequency shifts, energy intensity of specific frequency bands (such as fundamental frequency and second harmonic), spectral centroid, zero-crossing rate, and Mel-frequency cepstral coefficients can be extracted, collectively constituting the characteristic acoustic fingerprint. The essence of this fingerprint is a set of information that uniquely and stably reflects the current physical state of the interaction after most random noise has been removed.

[0076] Based on the characteristic acoustic fingerprint, the current tissue coupling state between the physiotherapy device and the human body is inferred;

[0077] Specifically, coupling state refers to the physical characteristics and quality of the interface between the therapeutic end and human tissue. It includes not only the tightness of contact (e.g., ideal fit, gaps, uneven pressure), but in more preferred embodiments, it may also include the tissue type characteristics of the treatment area (e.g., whether it is applied to a muscular or bony prominence). The inference process is a pattern recognition problem. The system uses a pre-established model or database to map the real-time extracted acoustic fingerprint features to various known coupling states. For example, it uses machine learning classifiers (e.g., Support Vector Machines (SVM), Convolutional Neural Networks (CNN)) or performs similarity matching with preset templates to ultimately output a judgment result for the current state, such as "ideal coupling," "poor coupling - gaps," "poor coupling - insufficient coupling agent," or "applying to adipose tissue," "applying to muscle tissue," etc.

[0078] In response to the tissue coupling state, a first adjustment command is generated for at least one output operating parameter of the temperature-controlled instrument to optimize the energy transfer efficiency between the therapeutic end and the human tissue.

[0079] Specifically, the system generates a first adjustment command based on the inferred coupling state. The goal of this command is to optimize energy transfer efficiency. For example, when the state is inferred to be "poor coupling," the command might increase the output power to compensate for energy loss at the contact interface, ensuring that the energy dose reaching deep tissues remains within the treatment window. Conversely, when the state is "ideal coupling," the command might maintain or slightly reduce the power to the optimal treatment value to avoid energy waste or potential overheating risks. In addition to power, "output operating parameters" may also include the frequency, amplitude, modulation mode of the drive signal, or the duration of the treatment head's operation. This command is sent to the main controller of the temperature and flux unit, thereby adjusting the transducer's drive circuitry in real time, forming a complete adaptive control closed loop.

[0080] This implementation method can automatically adapt to different users, different body parts, and real-time contact changes, thereby ensuring the precise delivery of therapeutic energy. This directly leads to improved stability of treatment effects (avoiding treatment ineffectiveness due to poor coupling) and enhanced safety (avoiding burns caused by energy buildup). Its scientific principle is based on acoustic impedance and structural dynamics. Different coupling states and tissue types have different acoustic impedances. When the vibration wave from the therapeutic end is transmitted to the interface, impedance mismatch causes changes in the wave reflection and transmission ratio. These changes immediately alter the vibration characteristics of the therapeutic end itself (i.e., change its acoustic fingerprint).

[0081] In a preferred embodiment, processing the mixed acoustic signal to separate and extract the characteristic acoustic fingerprint includes the following steps:

[0082] Under predetermined conditions where the physiotherapy end of the temperature-controlled instrument is unloaded and in operation, a baseline acoustic fingerprint is collected and established, wherein the baseline acoustic fingerprint characterizes the vibration noise;

[0083] Real-time mixed acoustic signals are acquired during normal treatment;

[0084] Specifically, "no-load" refers to the state where the treatment device is not in contact with the human body or any load, typically operating in the air. "Predetermined conditions" refers to the device operating under several standard, fixed output parameters (such as rated power and reference frequency). In this state, the signals collected by the acoustic sensors almost entirely originate from the device's own vibrations, i.e., vibration noise. The system filters, performs time-frequency analysis, and extracts features from this signal to establish a feature vector, which is the baseline acoustic fingerprint. This fingerprint is equivalent to the device's acoustic identity card, uniquely characterizing the device's inherent vibration characteristics under no-load conditions. This baseline fingerprint is stably stored in the device's memory as a reference for subsequent signal processing. During normal treatment, real-time mixed acoustic signals are acquired: this step is performed synchronously with the acquisition steps described above. When the device begins treating the user, a mixed acoustic signal containing noise and the applied signal is continuously acquired.

[0085] The real-time mixed acoustic signal is compared with the baseline acoustic fingerprint, and the characteristic acoustic fingerprint generated by the interaction between the therapeutic end and human tissue is separated by calculating the difference between the two in a predetermined frequency domain.

[0086] Specifically, both the real-time mixed acoustic signal and the original signal corresponding to the baseline acoustic fingerprint are transformed to the frequency domain using Fourier transform to obtain their respective spectra. The difference between the two spectra is calculated in one or more "predetermined frequency domains" (typically the device's fundamental frequency, harmonics, and characteristic frequency bands sensitive to load changes). The most direct method is spectral subtraction: the active signal spectrum ≈ the mixed signal spectrum - the baseline noise spectrum. From the calculated difference spectrum (i.e., the purified "active signal" spectrum), features (such as the energy and peak values ​​of the difference spectrum) are extracted again to ultimately form a characteristic acoustic fingerprint that truly represents tissue interactions for state inference.

[0087] This implementation takes into account that under stable operating conditions, the noise of the device itself is relatively stable and repeatable. The tissue interaction signal is "effective information" superimposed on the noise. Through baseline calibration, a "negative sample" of noise is essentially established in the system, thereby enabling the extraction of effective information from the mixed signal to the maximum extent, and significantly improving the signal-to-noise ratio.

[0088] In a preferred embodiment, inferring the current coupling state between the therapeutic device and human tissue based on the characteristic acoustic fingerprint includes the following steps:

[0089] The characteristic acoustic fingerprint is matched with a preset acoustic coupling template database to obtain the target physical coupling scene corresponding to the characteristic acoustic fingerprint;

[0090] The preset acoustic coupling template database includes multiple preset acoustic coupling templates and physical coupling scenarios corresponding to each preset acoustic coupling template. The physical coupling scenarios include ideal coupling, poor coupling, and coupling across different tissue types.

[0091] The target physical coupling scenario is used as the inferred organizational coupling state.

[0092] Specifically, the preset acoustic coupling template database is a database pre-built before shipment or through initial learning. Each preset acoustic coupling template in the database is essentially a standard "feature acoustic fingerprint" collected and stored under a specific known physical coupling scenario. Physical coupling scenarios include, but are not limited to: ideal coupling (perfect fit between the treatment head and skin, sufficient coupling agent), poor coupling (e.g., significant air gaps, dried coupling agent), and crossing different tissue types (e.g., transitioning from the abdomen with a thick fat layer to a thinner, near-bone area like the anterior tibia). The matching process involves calculating the similarity between the real-time extracted fingerprint and all template fingerprints in the database (e.g., calculating Euclidean distance, cosine similarity, or using a more complex Dynamic Time Warping-DTW algorithm) to find the most similar template. Through this matching, the system finds the best-matching template. This template, when pre-stored in the database, has already been labeled with its corresponding physical scenario. For example, a template collected in a laboratory under ideal coupling conditions is labeled "ideal coupling." Therefore, when the real-time fingerprint is most similar to this template, the system infers that the current target physical coupling scenario is ideal coupling. Ultimately, the system outputs no longer an abstract feature vector or similarity score, but a conclusive state description with clear physical and clinical significance, such as currently being in a malcoupled state or currently acting on a skeletal site. This conclusion will be directly used to generate the first adjustment instruction.

[0093] In a preferred embodiment, the preset acoustic coupling template database is a multidimensional dynamic database; wherein each stored preset acoustic coupling template is associated with at least one of the following dimension labels:

[0094] User identification dimension, used to distinguish the individual organizational characteristics of different users;

[0095] Body part dimension, used to distinguish different anatomical sites where the physical therapy device is applied;

[0096] The treatment time sequence dimension is used to characterize the different treatment time stages from the start to the end of a single treatment session.

[0097] The step of matching the characteristic acoustic fingerprint with a preset acoustic coupling template database to obtain the target physical coupling scene corresponding to the characteristic acoustic fingerprint includes the following steps:

[0098] Based on the current user identifier, and / or the identified body part, and / or the current treatment time stage, a target template subset is selected from the preset acoustic coupling template database;

[0099] The real-time acquired acoustic fingerprint features are matched with preset acoustic coupling templates in the target template subset to determine the target physical coupling scenario.

[0100] Specifically, multidimensionality refers to the fact that each template record in the database is accompanied by a set of metadata tags describing its collection background, while dynamism means that the database is not fixed after leaving the factory, but can be expanded and updated in subsequent use. Physically, the database can be a relational database (such as SQLite) or a NoSQL database. Each template is treated as a record, and its fields, in addition to the template data itself (i.e., feature vector), also include the following dimension tags. Specifically associated dimension tags include: User Identification Dimension: This tag is used to distinguish the individual organizational characteristics of different users. Upon first use, the system can guide the user to create a profile (e.g., User001). Subsequently, all templates collected and learned under this user (including their personal scenarios such as "ideal coupling" and "acting on the shoulder") will be tagged "User001". This solves the problem of different acoustic responses caused by differences in individual skin thickness, subcutaneous fat content, muscle density, etc. Body Part Dimension: This tag is used to distinguish different anatomical sites where the physiotherapy device is applied. The system can automatically identify treatment areas by user manual selection on the app (e.g., "shoulder," "waist," "calf") or through integrated simple position sensors (e.g., inertial measurement unit, IMU). Different body parts exhibit distinct acoustic fingerprints due to differences in underlying tissue structures and bone distances. Treatment Sequence Dimension: This label characterizes the different treatment time stages from the start to the end of a single treatment session. For example, a 20-minute treatment session can be divided into "initial stage (0-5 minutes)," "intermediate stage (5-15 minutes)," and "final stage (15-20 minutes)." Due to potential microcirculatory changes and slight swelling in tissues caused by thermal effects during treatment, their acoustic properties will slowly drift over time. This sequence dimension allows the system to use slightly different "ideal coupling" templates at different stages to accommodate these physiological dynamics. When state inference is required, the system first performs context awareness. It acquires current contextual information: which user is currently logged in (user identifier), which body part the treatment head is placed on (body part), and how long the treatment has been ongoing (treatment sequence). It then sends a query to the database, such as: "Find all acoustic coupling templates labeled User001, waist, mid-stage." The database returns a significantly reduced set of templates highly relevant to the current context, known as the "target template subset." Matching algorithms (such as calculating Euclidean distance) now only need to be performed within this refined "target template subset." The advantage of this is that it avoids meaningless cross-domain comparisons between the acoustic fingerprints of user A's waist and user B's calves, greatly improving matching accuracy and speed, and significantly reducing the risk of false positives.

[0101] In a preferred embodiment, after matching the real-time acquired acoustic fingerprint with a preset acoustic coupling template in the target template subset, the following step is further included:

[0102] Calculate the confidence score between the characteristic acoustic fingerprint and the best matching template in the target template subset;

[0103] Determining the target physical coupling scenario includes the following steps:

[0104] If the confidence score is higher than or equal to the preset confidence threshold, then the physical coupling scenario corresponding to the best matching template is determined as the target physical coupling scenario.

[0105] Specifically, after finding the best-matching template, the system calculates a confidence score. This score is a quantitative value used to measure the similarity between the real-time fingerprint and the best-matching template. The calculation method can vary: Distance-based scoring: If Euclidean distance is used for matching, the confidence score can be the reciprocal or a negative value of the distance (i.e., the smaller the distance, the higher the confidence). For normalization, it can be mapped to a 0-1 range. Similarity-based scoring: If cosine similarity is used, the similarity value itself (a number between 0 and 1) can be directly used as the confidence score. Statistical scoring: The difference between the mean and variance of the real-time fingerprint and the best-matching template can be considered, or the difference between it and the next best-matching template in the target subset can be calculated. If the best-matching template is significantly ahead, the confidence is high; if the scores of the first few templates are similar, the confidence is low. The system presets a "preset confidence threshold" (e.g., 0.8). This is an empirical value, set after testing on a large amount of experimental data. When the calculated confidence score is greater than or equal to this threshold, it indicates that the matching is very reliable. The system can confidently take the scenario corresponding to the best matching template (such as "ideal coupling") as the final inference result (i.e. "target physical coupling scenario") and generate adjustment instructions based on it.

[0106] In a preferred embodiment, after calculating the confidence score between the characteristic acoustic fingerprint and the best-matching template in the target template subset, the method further includes the following steps:

[0107] If the confidence score is lower than the preset confidence threshold, the characteristic acoustic fingerprint is determined to be candidate abnormal data;

[0108] A continuous observation window is started for the candidate abnormal data, and the frequency and stability of similar abnormal data with similar characteristics are monitored within the observation window;

[0109] Based on the monitoring results, perform one of the following operations:

[0110] If, within the observation window, abnormal data with similar characteristics appear frequently and stably, it is determined that they constitute a new effective physical coupling scenario, and a new acoustic coupling template is generated based on the similar abnormal data and stored in the preset acoustic coupling template database.

[0111] If the abnormal data appears only at low frequency or transiently within the observation window, it is determined to be interference noise, and a preset safety control strategy is triggered.

[0112] Specifically, when the confidence score falls below a threshold (e.g., 0.8), it indicates that the current acoustic fingerprint cannot be reliably interpreted by any known template in the database. Instead of forcing an unreliable inference, the system marks it as "candidate anomalous data," avoiding hasty action based on a single anomaly. It opens a continuous observation window, for example, lasting 30 seconds or the next 10 sampling cycles. Within this window, the system continuously monitors new acoustic fingerprints. More importantly, it uses clustering algorithms (e.g., K-means) or simple distance metrics to determine whether the newly emerging anomalous data shares "similar characteristics" with the initial candidate anomalous data. Simultaneously, it records the "frequency" (e.g., 8 times within 10 seconds) and "stability" (the degree of clustering of their feature vectors in the feature space, whether dispersed or concentrated) of these similar anomalous data. After the observation window ends, the system makes a final judgment based on the monitoring results: high frequency (e.g., exceeding 70% of the total samples within the window) and stability (feature vector clustering radius less than a certain threshold) indicate that this is not random noise, but a recurring, regular new pattern. This typically means the user is using the device in a way that the device hasn't learned before but is effective (such as in a specific body position or on a body part that hasn't been recorded). The system determines this to be a "new effective physical coupling scenario." It automatically calculates a new feature vector center as a "new acoustic coupling template" based on these stable clusters of anomalous data and stores it in the database. At this point, the system can prompt the user: "New usage pattern detected, learned and saved." Afterward, the system has the ability to recognize this new scenario. If, within the observation window, anomalous data only appears occasionally (low frequency), or its characteristics are erratic (unstable), it is likely due to external, accidental interference noise, such as the user suddenly tapping the device, the therapeutic end briefly leaving the skin and then returning, or strong environmental electromagnetic interference. The system determines this to be invalid interference, does not learn from it, and triggers a preset safety control strategy. This safety strategy might include automatically reducing the power to a safe default level and possibly alerting the user via sound or indicator light: "Signal is interfered with, please check device placement," until the signal returns to normal.

[0113] In a preferred embodiment, the following steps are also included:

[0114] Based on the characteristic acoustic fingerprint, the tissue temperature of the treatment area is calculated in real time.

[0115] The tissue temperature is compared with a preset target temperature range;

[0116] In response to the tissue temperature deviating from the preset target temperature range, a second adjustment command is generated for the operating parameters of the temperature transmitter to dynamically adjust the energy output and bring the tissue temperature back to the preset target temperature range.

[0117] Specifically, the tissue temperature is calculated by analyzing the mapping relationship between the frequency domain characteristics of the characteristic acoustic fingerprint and temperature-related acoustic parameters; the temperature-related acoustic parameters include sound velocity variation, sound attenuation coefficient, or resonant frequency shift.

[0118] Specifically, based on characteristic acoustic fingerprints, the tissue temperature of the treatment area is calculated in real time, relying on a pre-established physical mapping model. This model describes the quantitative relationship between acoustic features and tissue temperature. Data foundation: Under laboratory conditions and at known and precisely controlled tissue temperatures (calibrated via invasive thermocouples), corresponding "characteristic acoustic fingerprints" are collected, thus establishing a dataset of {temperature T, acoustic feature F}. Model establishment: Based on this dataset, a mapping function is trained using linear / nonlinear regression or a neural network: T = f(F). Here, the input F is the "temperature-related acoustic parameter" extracted from the characteristic acoustic fingerprint. Parameter extraction and calculation: During real-time treatment, the system continuously calculates these acoustic parameters from the characteristic acoustic fingerprint and substitutes them into the mapping function f(F) to calculate the current tissue temperature estimate in real time. Specifically, temperature-related acoustic parameters include: Sound velocity change: The sound velocity in biological tissues typically shows a slight positive correlation with increasing temperature (approximately 0.1%-0.2% / ℃). This change can be indirectly reflected by measuring the flight time of sound waves in tissue or the shift in the system's resonant frequency. For example, for a resonant system of fixed size, an increase in the speed of sound will lead to an increase in the resonant frequency. Sound attenuation coefficient: The absorption coefficient of tissue for sound waves generally increases with increasing temperature. This is reflected in the received echo or transmitted signal, where energy attenuation in specific frequency bands (especially high-frequency components) is amplified. Temperature changes can be inferred by monitoring changes in the ratio of high-frequency to low-frequency energy in the signal spectrum. Resonant frequency shift: The therapeutic end and the contacting tissue together constitute a complex vibration system. When changes in tissue temperature cause a change in its elastic modulus (stiffness), the natural resonant frequency of the entire system will drift. This drift is an extremely sensitive indicator for monitoring changes in temperature / mechanical state. The system has a preset target temperature range, such as 41°C to 45°C, which is considered a safe and effective treatment window. The real-time calculated tissue temperature is compared to this range to determine whether it is too low, normal, or too high. When the temperature deviates from the target range, the system generates a "second adjustment command." For example, if the temperature is below the lower limit, the second command is to increase the output power to raise the tissue temperature. If the temperature exceeds the upper limit, the second instruction is to reduce the output power or suspend the output to prevent tissue overheating damage. The core objective of the second adjustment instruction is thermal management, ensuring that the treatment is within the optimal thermobiological effect range.

[0119] This implementation method achieves non-invasive, real-time monitoring of deep tissue temperature and temperature-based safe closed-loop control. It fundamentally solves the problems of overheating risks or insufficient treatment caused by traditional temperature monitoring devices' inability to sense the true thermal state inside the tissue and their reliance on experience-based parameter settings. This greatly improves the safety of the device and the controllability of the treatment.

[0120] In a preferred embodiment, the following steps are also included:

[0121] Based on the tissue coupling state and the tissue temperature, a corresponding priority decision rule is obtained;

[0122] Based on the aforementioned priority decision rules, target adjustment instructions are generated to synergistically optimize energy transfer efficiency and thermal management objectives.

[0123] The priority decision rules include:

[0124] When the absolute value of the difference between the tissue temperature and the preset safety threshold is less than the first preset threshold, the second adjustment command is selected as the target adjustment command;

[0125] When the absolute value of the difference between the tissue temperature and the preset safety threshold is greater than or equal to the first preset threshold, the first adjustment instruction is selected as the target adjustment instruction.

[0126] Specifically, the system internally establishes a set of logical rules, namely priority decision rules. The core concept of these rules is: safety is the top priority. The rules are designed to be highly sensitive to temperature conditions. Specifically, the rules include: when the absolute value of the difference between the temperature and a preset safety threshold is less than a first preset threshold, a second adjustment command is selected as the target adjustment command. The preset safety threshold is typically the upper limit of the target temperature range (e.g., 45°C). The first preset threshold is a safety margin, for example, 2°C. This means that when the real-time temperature reaches or is very close to (e.g., 43°C and above) the safety threshold, the system enters a "high-temperature alarm state." In this state, regardless of the coupling state (even if poorly coupled), thermal safety becomes the absolute priority. The system will ignore the first adjustment command aimed at optimizing efficiency (e.g., a request for increased power due to poor coupling) and directly adopt the second adjustment command requiring cooling or maintenance as the final "target adjustment command" to be executed. "When the absolute value of the difference between the tissue temperature and the preset safety threshold is greater than or equal to the first preset threshold, the first adjustment command is selected as the target adjustment command." When the real-time temperature is far from the safety threshold (e.g., below 43°C) and within the safe operating range, the system considers there to be sufficient room to optimize treatment efficiency. Therefore, it will adopt the first adjustment command based on the coupling state as the "target adjustment command." For example, if coupling is poor, the power will be increased to ensure energy effectively penetrates the tissue.

[0127] This implementation achieves an intelligent trade-off and balance between safety and efficiency. It actively optimizes performance within safety boundaries and decisively safeguards safety when safety red lines are crossed, resulting in optimal and most rational overall control behavior.

[0128] In a preferred embodiment, after processing the mixed acoustic signal, the method further includes the following step:

[0129] From the characteristic acoustic fingerprint, acoustic elastic parameters for characterizing the physiological state of the tissue are extracted; the acoustic elastic parameters are obtained by analyzing the energy distribution changes or resonant peak shifts in specific frequency bands of the characteristic acoustic fingerprint.

[0130] Based on the changing trend of the acoustic elasticity parameters over the treatment time series, the physiological response intensity of human tissue during the treatment process can be obtained in real time.

[0131] After generating a first adjustment command for at least one output operating parameter of the temperature transmitter in response to the tissue coupling state, the method further includes the following steps:

[0132] Based on the intensity of the physiological response, the first adjustment instruction is modified to generate an optimized first adjustment instruction;

[0133] The correction rule is as follows:

[0134] When the physiological response intensity is lower than the expected response threshold, the adjustment range of the energy output parameter in the first adjustment command is increased;

[0135] When the physiological response intensity is higher than the expected response threshold, the adjustment range of the energy output parameter in the first adjustment command is reduced or maintained.

[0136] Specifically, acoustic elasticity parameters are indicators reflecting the stiffness (elastic modulus) of tissue. They are extracted as follows: by analyzing the energy distribution changes in specific frequency bands within the characteristic acoustic fingerprint. When tissue softens, its absorption of low-frequency vibrations weakens, while its absorption of high-frequency vibrations may increase, leading to a shift in the energy distribution of the signal spectrum (e.g., the spectral centroid). Analyzing the shift in the resonant peak value of specific frequency bands within the characteristic acoustic fingerprint is a more sensitive method. The resonant frequency of the vibration system formed by the therapeutic end and the tissue is directly related to the tissue's equivalent stiffness. During treatment, if the tissue softens due to thermal effects or microstructural changes (reduced elastic modulus), the system's resonant frequency will shift towards lower frequencies. Monitoring this shift quantitatively reflects the change in tissue elasticity. The system does not look at the elasticity parameter value at a single point in time, but rather analyzes its "trend" over time. For example, in the early stages of treatment, as the tissue begins to relax due to heat and mechanical stimulation, the resonant frequency will continuously decrease; at this time, the "physiological response intensity" is considered strong. As treatment progresses, the rate of decrease in frequency slows down and even plateaus, indicating that the tissue has adapted to the intensity of treatment and the response intensity has weakened.

[0137] Specifically, when the physiological response intensity is lower than the expected response threshold, the adjustment range of the energy output parameter in the first adjustment instruction is increased: this indicates that the tissue is insensitive to the current dose, and the treatment may have entered a plateau phase. The system will then make enhancement corrections based on the original first adjustment instruction. For example, if the original instruction suggested maintaining power due to ideal coupling, the optimized instruction may change to slightly increasing power to re-excite the tissue's physiological response. When the physiological response intensity is higher than the expected response threshold, the adjustment range of the energy output parameter in the first adjustment instruction is decreased or maintained: this indicates that the tissue responds strongly to treatment and is very sensitive. The system will make conservative corrections. For example, if the original instruction suggested "significantly increasing power" due to "poor coupling," but given the high physiological response, the optimized instruction may only "slightly increase power" or "temporarily not increase power" to avoid overstimulation.

[0138] This approach represents a significant advancement from "physical state-based regulation" to "physiological efficacy-based regulation." It enables the thermometer to dynamically adjust based on each individual's real-time response to each treatment, resolving the core issue that fixed parameters cannot address individual differences and dynamic changes in treatment. This maximizes therapeutic efficacy and minimizes ineffective treatment.

[0139] Secondly, this application proposes an adaptive control system for a temperature flow meter based on acoustic signal feedback, used to implement the adaptive control method for a temperature flow meter based on acoustic signal feedback as described above, including:

[0140] The acquisition module is configured to acquire mixed acoustic signals during the operation of the thermal oscillator via an acoustic sensor integrated inside the physiotherapy end of the thermal oscillator; the mixed acoustic signals include the vibration noise of the device itself and the interaction signals with human tissue.

[0141] An extraction module is configured to process the mixed acoustic signal to separate and extract characteristic acoustic fingerprints;

[0142] A judgment module is configured to infer the current coupling state between the physiotherapy terminal and human tissue based on the characteristic acoustic fingerprint.

[0143] A control module configured to generate a first adjustment command for at least one output operating parameter of the temperature transducer in response to the tissue coupling state, so as to optimize the energy transfer efficiency between the therapy end and the human tissue.

[0144] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. The above descriptions are only preferred embodiments of this application. It should be noted that due to the limitations of written expression, while there are objectively infinite specific structures, those skilled in the art can make several improvements, modifications, or changes without departing from the principles of this invention, and can also combine the above technical features in an appropriate manner. These improvements, modifications, changes, or combinations, or the direct application of the inventive concept and technical solution to other situations without modification, should all be considered within the scope of protection of this application.

Claims

1. A temperature and ventilation instrument adaptive control system based on acoustic signal feedback, characterized in that, include: The acquisition module is configured to acquire mixed acoustic signals during the operation of the thermal oscillator via an acoustic sensor integrated inside the physiotherapy end of the thermal oscillator; the mixed acoustic signals include the vibration noise of the device itself and the interaction signals with human tissue. An extraction module is configured to process the mixed acoustic signal to separate and extract characteristic acoustic fingerprints; A judgment module is configured to infer the current coupling state between the physiotherapy terminal and human tissue based on the characteristic acoustic fingerprint. A control module configured to generate a first adjustment command for at least one output operating parameter of the temperature transducer in response to the tissue coupling state, so as to optimize the energy transfer efficiency between the therapy end and the human tissue. The control method of the control system includes the following steps: The acoustic sensors integrated into the physiotherapy end of the thermometer collect mixed acoustic signals during the operation of the thermometer; the mixed acoustic signals include the vibration noise of the device itself and the interaction signals with human tissue. The mixed acoustic signals are processed to separate and extract characteristic acoustic fingerprints; Based on the characteristic acoustic fingerprint, the current tissue coupling state between the physiotherapy device and the human body is inferred; In response to the tissue coupling state, a first adjustment command is generated for at least one output operating parameter of the temperature transducer to optimize the energy transfer efficiency between the therapy end and the human tissue. After processing the mixed acoustic signal, the method further includes the following steps: From the characteristic acoustic fingerprint, acoustic elastic parameters for characterizing the physiological state of the tissue are extracted; the acoustic elastic parameters are obtained by analyzing the energy distribution changes or resonant peak shifts in specific frequency bands of the characteristic acoustic fingerprint. Based on the changing trend of the acoustic elasticity parameters over the treatment time series, the physiological response intensity of human tissue during the treatment process can be obtained in real time. After generating a first adjustment command for at least one output operating parameter of the temperature transmitter in response to the tissue coupling state, the method further includes the following steps: Based on the intensity of the physiological response, the first adjustment instruction is modified to generate an optimized first adjustment instruction; The revised rules are as follows: When the physiological response intensity is lower than the expected response threshold, the adjustment range of the energy output parameter in the first adjustment command is increased; When the physiological response intensity is higher than the expected response threshold, the adjustment range of the energy output parameter in the first adjustment command is reduced or maintained.

2. The adaptive control system for a temperature and flux meter based on acoustic signal feedback according to claim 1, characterized in that: The process of processing the mixed acoustic signal to separate and extract the characteristic acoustic fingerprint includes the following steps: Under predetermined conditions where the physiotherapy end of the temperature-controlled instrument is unloaded and in operation, a baseline acoustic fingerprint is collected and established, wherein the baseline acoustic fingerprint characterizes the vibration noise; Real-time mixed acoustic signals are acquired during normal treatment; The real-time mixed acoustic signal is compared with the baseline acoustic fingerprint, and the characteristic acoustic fingerprint generated by the interaction between the therapeutic end and human tissue is separated by calculating the difference between the two in a predetermined frequency domain.

3. The adaptive control system for a temperature and flux meter based on acoustic signal feedback according to claim 1, characterized in that: The step of inferring the current coupling state between the therapeutic device and human tissue based on the characteristic acoustic fingerprint includes the following steps: The characteristic acoustic fingerprint is matched with a preset acoustic coupling template database to obtain the target physical coupling scene corresponding to the characteristic acoustic fingerprint; The preset acoustic coupling template database includes multiple preset acoustic coupling templates and physical coupling scenarios corresponding to each preset acoustic coupling template. The physical coupling scenarios include ideal coupling, poor coupling, and coupling across different tissue types. The target physical coupling scenario is used as the inferred organizational coupling state.

4. The adaptive control system for a temperature and flux meter based on acoustic signal feedback according to claim 3, characterized in that: The preset acoustic coupling template database is a multidimensional dynamic database; each preset acoustic coupling template stored therein is associated with at least one of the following dimension labels: User identification dimension, used to distinguish the individual organizational characteristics of different users; Body part dimension, used to distinguish different anatomical sites where the physical therapy device is applied; The treatment time sequence dimension is used to characterize the different treatment time stages from the start to the end of a single treatment session. The step of matching the characteristic acoustic fingerprint with a preset acoustic coupling template database to obtain the target physical coupling scene corresponding to the characteristic acoustic fingerprint includes the following steps: Based on the current user identifier, and / or the identified body part, and / or the current treatment time stage, a target template subset is selected from the preset acoustic coupling template database; The real-time acquired acoustic fingerprint features are matched with preset acoustic coupling templates in the target template subset to determine the target physical coupling scenario.

5. The adaptive control system for a temperature and flux meter based on acoustic signal feedback according to claim 4, characterized in that: After matching the real-time acquired acoustic fingerprint with the preset acoustic coupling templates in the target template subset, the following steps are also included: Calculate the confidence score between the characteristic acoustic fingerprint and the best matching template in the target template subset; Determining the target physical coupling scene includes the following steps: If the confidence score is higher than or equal to the preset confidence threshold, then the physical coupling scenario corresponding to the best matching template is determined as the target physical coupling scenario.

6. The adaptive control system for a temperature and flux meter based on acoustic signal feedback according to claim 5, characterized in that: After calculating the confidence score between the characteristic acoustic fingerprint and the best matching template in the target template subset, the method further includes the following steps: If the confidence score is lower than the preset confidence threshold, the characteristic acoustic fingerprint is determined to be candidate abnormal data; A continuous observation window is started for the candidate abnormal data, and the frequency and stability of similar abnormal data with similar characteristics are monitored within the observation window; Based on the monitoring results, perform one of the following operations: If, within the observation window, abnormal data with similar characteristics appear frequently and stably, it is determined that they constitute a new effective physical coupling scenario, and a new acoustic coupling template is generated based on the similar abnormal data and stored in the preset acoustic coupling template database. If the abnormal data appears only at low frequency or transiently within the observation window, it is determined to be interference noise, and a preset safety control strategy is triggered.

7. The adaptive control system for a temperature and flux meter based on acoustic signal feedback according to claim 1, characterized in that: It also includes the following steps: Based on the characteristic acoustic fingerprint, the tissue temperature of the treatment area is calculated in real time. The tissue temperature is compared with a preset target temperature range; In response to the tissue temperature deviating from the preset target temperature range, a second adjustment command is generated for the operating parameters of the temperature transmitter to dynamically adjust the energy output and bring the tissue temperature back to the preset target temperature range. Specifically, the tissue temperature is calculated by analyzing the mapping relationship between the frequency domain characteristics of the characteristic acoustic fingerprint and temperature-related acoustic parameters; the temperature-related acoustic parameters include sound velocity variation, sound attenuation coefficient, or resonant frequency shift.

8. The adaptive control system for a temperature and flux meter based on acoustic signal feedback according to claim 7, characterized in that: It also includes the following steps: Based on the tissue coupling state and the tissue temperature, a corresponding priority decision rule is obtained; Based on the aforementioned priority decision rules, target adjustment instructions are generated to synergistically optimize energy transfer efficiency and thermal management objectives. The priority decision rules include: When the absolute value of the difference between the tissue temperature and the preset safety threshold is less than the first preset threshold, the second adjustment command is selected as the target adjustment command; When the absolute value of the difference between the tissue temperature and the preset safety threshold is greater than or equal to the first preset threshold, the first adjustment instruction is selected as the target adjustment instruction.

Images