Physiotherapy lamp irradiation adjusting system and method based on human body state perception

Abstract

The invention relates to the field of physiotherapy lamp irradiation adjustment, and discloses a physiotherapy lamp irradiation adjustment system and method based on human body state perception, and the method comprises the steps: collecting a behavior response signal, a physiological reaction signal and a posture stability signal of an irradiated human body in parallel, executing time synchronization, abnormity elimination and scale normalization processing, and constructing a physiotherapy state model; based on historical physiotherapy data and a current human body state, extracting failure constraint factors from the physiotherapy state model, and mapping the failure constraint factors into a constraint boundary parameter set; determining a state feasible region corresponding to the irradiation adjustment strategy, calibrating a state region exceeding the state feasible region as a failure risk region, and constructing a three-layer constraint space structure; continuously tracking the human body state evolution trajectory, and generating a combined adjustment instruction; and combining the determined state feasible region, dynamically switching the irradiation control mode, and synchronously updating the failure constraint factor and the state feasible region parameter to form an updated irradiation control state. The method has the advantage of improving the physical therapy adaptability.

Description

Technical Field

[0001] This invention relates to the field of physiotherapy lamp irradiation adjustment, specifically to a physiotherapy lamp irradiation adjustment system and method based on human body state perception. Background Technology

[0002] In existing physiotherapy lamp irradiation adjustment systems, the irradiation process is mainly controlled based on physical or physiological safety indicators such as human body surface temperature, irradiation time, output power, and some physiological parameters to avoid overheating or excessive irradiation that could harm the human body. However, in practical applications, especially among people recovering from surgery, those sensitive to pain, those with nervous tension, or those with weak psychological resilience, patients are prone to emotional fluctuations such as tension, anxiety, fear, or resistance during physiotherapy. These emotional fluctuations often cause subconscious avoidance behaviors, muscle contractions, changes in posture, or interruptions in physiotherapy, making it impossible for the originally set irradiation parameters to truly correspond to the actual irradiation state of the human body. This leads to unstable physiotherapy effects or even adverse reactions. In addition, existing physiotherapy lamp systems lack sufficient mechanisms for perceiving and quantifying the patient's emotional state, and have not established a regulatory correlation between emotional changes and irradiation parameters. Therefore, they cannot make real-time adjustments when the human body's physiological or psychological state changes, making the physiotherapy logic limited to physical safety constraints. This makes it difficult to adapt to the differences in psychological and neurological responses among different individuals, resulting in poor physiotherapy experience, large fluctuations in efficacy, and low patient acceptance. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a physiotherapy lamp irradiation adjustment system and method based on human body state perception, which has the advantage of improving the adaptability of physiotherapy and solves the problems mentioned in the background technology.

[0004] To achieve the aforementioned goal of improving the adaptability of physiotherapy, the present invention provides the following technical solution: a method for adjusting physiotherapy lamp irradiation based on human body state perception, comprising the following steps: Parallel acquisition of behavioral response signals, physiological reaction signals, and posture stability signals of the irradiated human body; time synchronization, anomaly removal, and scale normalization of the signals; and construction of a physiotherapy state model characterizing the evolution of the human body's irradiation response. Based on historical physiotherapy data and current human body condition, failure constraint factors reflecting the risk of physiotherapy failure, the trend of radiation deviation, and the probability of human body resistance reaction are extracted from the physiotherapy state model and mapped to the set of constraint boundary parameters in the physiotherapy state model. Based on the set of constraint boundary parameters, the state feasible region corresponding to the irradiation adjustment strategy is determined, and the state region that exceeds the state feasible region is marked as the failure risk zone, thus constructing a three-layer constraint space structure that limits the human state evolution path. Based on a three-layer constrained spatial structure, the system continuously tracks the trajectory of human body state evolution. When the trajectory deviates from the failure risk zone or when the human body shows an abnormal trend in subjective feelings, the system adjusts the weights of physiotherapy orientation and comfort orientation and executes comfort orientation adjustment to generate joint adjustment instructions. Based on the trajectory corrected by the joint adjustment command, and combined with the determined state feasible region, the irradiation control mode is dynamically switched, and the failure constraint factor and state feasible region parameters are updated synchronously during the mode switching process to form the updated irradiation control state.

[0005] Preferably, the process of constructing a physical therapy state model characterizing the evolution of the human body's irradiation response is as follows: Deploy multi-channel sensor acquisition nodes within the irradiation area, including motion sensors, heart rate and skin conductance sensors, posture detection devices, and ambient light intensity sensors; The signals from each sensor are synchronized with microsecond-level timestamps, and signal consistency is ensured through filtering, outlier removal, and normalization algorithms. Based on the processed multi-source signals, a multi-dimensional temporal feature extraction and weighted fusion method is used to generate a feature sequence that reflects the evolution of the human body's response to irradiation, and a physiotherapy state model that characterizes the evolution of the human body's response to irradiation is constructed accordingly.

[0006] Preferably, the process of extracting failure constraint factors reflecting the risk of physiotherapy failure, the trend of radiation deviation, and the probability of human resistance from the physiotherapy state model is as follows: Input historical data and the current human body status into the constructed physiotherapy status model; By using statistical modeling and Bayesian update methods, the probability distribution and trend estimate of human state deviation under different irradiation conditions are generated. The probability distribution and trend are used to construct a multidimensional state representation of the physiotherapy state model. As input, multi-task regression analysis and reinforcement learning algorithms are applied to predict potential resistance responses and generate quantitative failure constraint factors.

[0007] Preferably, the process of mapping to the set of constraint boundary parameters in the physiotherapy state model is as follows: Based on the multidimensional state characteristics of the human body, the failure constraint factors are divided into three layers of constraint space according to state sensitivity and time dependence, including low risk, adjustable risk and high risk areas. Define specific numerical ranges and evolution weights for each boundary layer; A dynamic mapping algorithm is executed for each boundary parameter to weight and fuse the probability value, weight, and trend parameter of the failure constraint factor with the current state and historical state evolution trajectory of the human body, generating a set of real-time adjustable constraint boundary parameters.

[0008] Preferably, the process of determining the feasible region of the irradiation adjustment strategy is as follows: By taking the set of constraint boundary parameters as input and combining them with the current multidimensional state characteristics of the human body, a multidimensional illumination parameter mapping matrix is ​​constructed in the three-layer constraint space. The parameters of light intensity, wavelength, illumination angle, and duration are mapped to low-risk, adjustable-risk, and high-risk zones. Weighted fusion calculations are performed on each parameter dimension and its corresponding risk zone, incorporating the probability values, evolution trends, and human state sensitivity of the constraint boundary parameters into the weighting factors. The upper and lower limits, allowable step size, and safety margin of each parameter are calculated by calling a multi-objective optimization algorithm to form a structured feasible state region.

[0009] Preferably, the process of constructing a three-layer constrained spatial structure that limits the evolutionary path of the human body state is as follows: The combined illumination parameters outside the feasible state domain are calibrated as the failure risk zone, and low-risk, adjustable-risk, and high-risk parameter levels are established in the three-layer constraint space. To address the multidimensional state characteristics of the human body, each dimension is mapped to a three-layer constraint space, and boundary functions, variation ranges, and evolution weights are defined for each layer. Define legal transition rules for state evolution paths in the constraint space, including the allowed step size for increasing or decreasing state dimensions, the direction of parameter combination changes, and cross-layer transition conditions; The boundary parameters and migration rules are dynamically updated, and weighted corrections are performed based on real-time multi-source signals to form a complete three-layer constraint space structure.

[0010] Preferably, the process of continuously tracking the human body's state evolution trajectory based on a three-layer constrained spatial structure is as follows: Collect and record multimodal response data of the human body during irradiation, including behavioral response, physiological reaction and posture stability signals, and map each state dimension to a three-layer constraint space structure; Calculate the numerical position, deviation, and cross-layer migration risk index of each state dimension in the constraint space to generate a multi-dimensional state deviation matrix; Based on the legal migration rules of the evolution path defined in the three-layer constraint space, and combined with the multi-dimensional state deviation matrix, the human state trajectory is gradually tracked and analyzed in time to identify the trajectory sequence of potential deviation failure risk areas. Data annotation and indexing are performed on the trajectory sequence to generate a trajectory mapping table.

[0011] Preferably, the process of generating joint control instructions is as follows: Using the multidimensional state deviation matrix and trajectory mapping table as input, the deviation magnitude, direction and cross-layer migration risk of each state dimension in the three-layer constraint space are quantified. The quantification results are weighted and combined with the historical evolution information of the trajectory mapping table to construct a deviation evaluation characterization; Based on the deviation assessment characterization, the irradiation guidance control parameters and the weights of each state dimension are calculated according to the preset weight adjustment algorithm, and a joint adjustment command is generated.

[0012] Preferably, the process for forming the updated irradiation control state is as follows: Based on the joint adjustment instructions and the current status, determine the applicable irradiation mode and corresponding parameter combination; The failure constraint factor and state feasible domain parameters are calculated and updated in real time, and the current human state is mapped to the feasible space of the selected mode. Record the state mapping, irradiation parameter execution, and constraint parameter update during the mode switching process, and output the updated irradiation control state, including irradiation mode, parameter configuration, and state feasible domain mapping information.

[0013] A therapeutic lamp irradiation adjustment system based on human body state perception includes: State modeling module: Parallel acquisition of signals such as human behavior and heart rate, synchronization, anomaly removal and normalization, and construction of human irradiation response evolution characteristic model; Constraint Extraction Module: Based on historical physiotherapy data and the current state, it extracts failure constraint factors such as failure risk, deviation trend, and resistance probability from the physiotherapy state model and maps them into a set of constraint boundary parameters; Feasible domain delineation module: Based on the set of constraint boundary parameters, determine the state feasible domain of the illumination strategy, mark the region outside the feasible domain as the failure risk zone, and construct a three-layer constraint space to limit the state evolution path; Trajectory Analysis Module: Continuously tracks the evolution trajectory of the human body in the three-layer constraint space, and adjusts the irradiation weight to generate joint adjustment instructions when the body deviates from the failure risk zone or an abnormal trend occurs; Mode Management Module: Based on the human body state trajectory adjusted by the joint adjustment command, switch the irradiation mode and update the failure constraint factor and state feasible domain, and generate the updated irradiation control state.

[0014] Compared with the prior art, the present invention provides a physiotherapy lamp irradiation adjustment system and method based on human body state perception, which has the following beneficial effects: This invention acquires human behavioral response signals, physiological reaction signals, and posture stability signals in parallel, and performs time synchronization, anomaly removal, and scale normalization on multi-source signals to construct a physiotherapy state model characterizing the evolution of human irradiation response. This achieves comprehensive perception of the human irradiation state. Based on this state model and historical physiotherapy data, failure constraint factors reflecting the risk of physiotherapy failure, irradiation deviation trends, and the probability of human resistance are extracted and mapped to a set of constraint boundary parameters. This enables the quantification and structured management of the risk of irradiation parameter adjustment. The set of constraint boundary parameters is used to determine the feasible domain of the irradiation strategy, and the state regions outside the feasible domain are marked as failure risk zones. A three-layer constraint space structure is constructed to limit the human state evolution path, achieving safety constraints and path guidance for the human state evolution process. During irradiation, the three-layer constraint space structure continuously tracks... The system tracks the evolution trajectory of the human body's state and, when the trajectory deviates from the failure risk zone or when subjective abnormalities occur, adjusts the weights of physiotherapy guidance and comfort guidance to generate joint adjustment commands. This enables dynamic optimization control of irradiation parameters. Based on the joint adjustment commands, the system adjusts the human body's state trajectory and switches irradiation control modes in conjunction with the state feasible domain. During mode switching, the system updates failure constraint factors and state feasible domain parameters in real time to form an updated irradiation control state. This achieves continuous regulation and multi-dimensional constraint management throughout the entire irradiation process. By introducing multimodal human body state perception, failure constraint factor quantification, a three-layer constraint space, and a joint adjustment mechanism, the system achieves safety, controllability, personalized adaptability, and dynamic optimization of the physiotherapy lamp irradiation process. This overcomes the shortcomings of existing technologies, such as static irradiation control, lack of human body state feedback, and multi-dimensional constraint management, providing a complete and effective technical solution for the precise adjustment of intelligent physiotherapy lamps. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the method of the present invention; Figure 2 This is a schematic diagram of the structure of the present invention. Detailed Implementation

[0016] 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.

[0017] Example 1: Please refer to Figure 1 As shown in the figure, the method for adjusting the irradiation of a physiotherapy lamp based on human body state perception according to an embodiment of the present invention includes the following steps: S1: Parallel acquisition of behavioral response signals, physiological reaction signals, and posture stability signals of the irradiated human body; time synchronization, anomaly removal, and scale normalization of the signals; construction of a physiotherapy state model characterizing the evolution of the human body's irradiation response.

[0018] The process of constructing a physical therapy state model representing the evolution of the human body's irradiation response in S1 is as follows: Deploy multi-channel sensor acquisition nodes within the irradiation area, including motion sensors, heart rate and skin conductance sensors, posture detection devices, and ambient light intensity sensors; During physiotherapy, the system collects real-time data on the patient's movements, heart rate, skin conductance, body posture, and ambient light intensity. Movement data is captured using an inertial measurement unit (IMU) to measure the triaxial acceleration and angular velocity of the arms, torso, and head, and the amplitude, velocity, and acceleration trends are calculated. Heart rate is measured using a photoplethysmography (PPG) sensor to obtain the heart rate interval per second and heart rate variability. Skin conductance sensors measure real-time changes in the electrical conductance of the palms or fingers, reflecting autonomic nervous activity. Posture data is obtained using a depth camera or infrared imaging system to acquire the angles and stability indicators of major joints, and to analyze micro-movements and support stability. Ambient light intensity sensors measure local lighting conditions and the actual illuminance of the physiotherapy lights, ensuring that changes in ambient light do not interfere with physiological signals. All signals are sampled at frequencies between 100Hz and 500Hz and include microsecond-level timestamps to ensure high-precision capture of rapid movements and emotional fluctuations.

[0019] The signals from each sensor are synchronized with microsecond-level timestamps, and signal consistency is ensured through filtering, outlier removal, and normalization algorithms. Multi-source signals are time-aligned according to microsecond-level timestamps to ensure the synchronization of each sampling point. Low-pass and band-stop filtering are applied to behavioral, posture, and physiological signals to remove 50 / 60Hz power frequency interference and high-frequency noise. An outlier elimination algorithm is used to eliminate instantaneous abrupt changes or sensor loss points, and linear interpolation is performed to compensate for short-term missing data. Each signal is normalized to the 0–1 range or z-score normalized to ensure that signals of different dimensions can be processed with equal weight in feature fusion.

[0020] Based on the processed multi-source signals, a multi-dimensional temporal feature extraction and weighted fusion method is used to generate a feature sequence that reflects the evolution of the human body's response to irradiation, and a physiotherapy state model that characterizes the evolution of the human body's response to irradiation is constructed accordingly. A sliding window analysis was performed on each preprocessed signal sequence to extract time-domain features, such as mean, variance, kurtosis, slope, and rate of change of maximum / minimum values, and frequency-domain features, such as power spectral density, dominant frequency components, and energy concentration. Behavioral and postural features reflected irradiation deviation trends and micromotion stability, while heart rate and skin conductance reflected autonomic nervous system activity and emotional fluctuations. Ambient light features were used to correct for the influence of irradiation errors on physiological signals. Features from different dimensions were then fused according to weights, which could be obtained based on clinical experience or prior data training. For example, heart rate and skin conductance had higher weights to reflect emotional state, postural changes and behavioral features had medium weights to detect postural deviations and irradiation instability, and ambient light changes had lower weights to compensate for light interference. After fusion, a multidimensional feature vector is generated for each time point, constructing a continuous feature sequence that fully reflects the dynamic response evolution of the human body to the irradiation of the therapy lamp. Based on the fused feature sequence, the multidimensional features at each time point are mapped to a therapy state vector, including the current irradiation state, emotional state indicators, and irradiation deviation risk estimation. The model is updated at each sampling time step and can be combined with historical sequences for trend prediction and short-term evolution analysis to generate the human body state evolution trajectory. The state model is directly used to drive the irradiation intensity, direction, and rhythm adjustment of the therapy lamp to achieve closed-loop regulation. It can be updated in real time in the edge computing unit or local controller to ensure low-latency and high-precision control of the human body's response. At the same time, a data interface can be provided to upload to a remote monitoring system for long-term efficacy analysis and optimization.

[0021] S2: Based on historical physiotherapy data and the current human body condition, extract failure constraint factors from the physiotherapy state model that reflect the risk of physiotherapy failure, the trend of radiation deviation, and the probability of human body resistance reaction, and map them into a set of constraint boundary parameters in the physiotherapy state model.

[0022] The process in S2 of extracting failure constraint factors reflecting the risk of physiotherapy failure, the trend of radiation deviation, and the probability of human resistance from the physiotherapy state model is as follows: Input historical data and the current human body status into the constructed physiotherapy status model; The patient's historical physiotherapy records and current human body status data are input into the constructed physiotherapy status model. The historical records include multi-dimensional signal sequences such as irradiation parameters of past treatments, patient behavioral responses, heart rate, skin conductance response, and postural stability indicators. The current human body status is a real-time physiotherapy status vector. Missing value imputation, outlier removal, and standardization are performed on the input data to ensure the consistency and reliability of the data in statistical analysis and machine learning calculations.

[0023] By using statistical modeling and Bayesian update methods, the probability distribution and trend estimate of human state deviation under different irradiation conditions are generated. The processed data are modeled using statistical methods, such as Gaussian mixture models or Markov processes, to calculate the probability distribution of patient state deviation under different irradiation intensities, directions, and rhythms. Combined with a Bayesian update mechanism, the deviation trend of the patient's current state relative to historical data is dynamically updated, and the deviation probability and confidence interval of each time step are output. This ensures that the risk of patient irradiation deviation under real-time irradiation can be quantified, and provides a quantitative basis for trajectory correction.

[0024] The probability distribution and trend are used to construct a multidimensional state representation of the physiotherapy state model. As input, multi-task regression analysis and reinforcement learning algorithms are applied to predict potential resistance responses and generate quantitative failure constraint factors. The probability of human irradiation deviation, trend changes, and irradiation parameters at each sampling time point are integrated into a multi-dimensional state vector, forming a continuous state sequence. This sequence is input into a multi-task regression model to jointly predict deviation trends and potential resistance behaviors under different irradiation conditions. The regression output includes irradiation deviation risk value, posture shift, and autonomic nervous activity indicators. The regression results are further input into a reinforcement learning algorithm to construct a state-action-reward mechanism, simulating the possible response paths of patients under different irradiation strategies. The optimal irradiation strategy is generated by optimizing the reward function, and key risk indicators and constraint parameters are extracted from it, including deviation upper limit, potential resistance probability threshold, and feasible irradiation angle and intensity range. Finally, these quantitative indicators are organized into failure constraint factors to guide irradiation control strategies and trajectory correction, ensuring that the human body's state remains within a safe, controllable, and comfortable range during treatment.

[0025] The process of mapping S2 to the set of constraint boundary parameters in the physiotherapy state model is as follows: Based on the multidimensional state characteristics of the human body, the failure constraint factors are divided into three layers of constraint space according to state sensitivity and time dependence, including low risk, adjustable risk and high risk areas. Based on the human body's multidimensional state vector at each time point, the corresponding failure constraint factors are analyzed according to state sensitivity, such as heart rate variability, postural deviation amplitude, and intensity of behavioral movements, as well as time dependence. By using preset thresholds or statistical quantiles based on historical data, the failure constraint factors are automatically classified into three-layer constraint spaces: low-risk, adjustable-risk, and high-risk areas. Each area corresponds to different adjustment priorities and intervention sensitivities, ensuring that the physiotherapy control system can dynamically select adjustment strategies for different risk levels.

[0026] Define specific numerical ranges and evolution weights for each boundary layer; Specific numerical boundaries and evolution weights are set for each constraint region. The numerical range can be calculated based on historical physiotherapy data. For example, small deviations are allowed in the low-risk zone, medium deviations are allowed in the adjustable-risk zone, and critical deviations that must be corrected in time correspond to the high-risk zone. The evolution weights are used to control the response speed when the state moves closer to or away from the boundary. The weights can be initially set through clinical experience and dynamically adjusted through an adaptive algorithm during system operation to reflect the relative influence of different risk levels on irradiation regulation.

[0027] A dynamic mapping algorithm is executed for each boundary parameter to weight and fuse the probability value, weight and trend parameter of the failure constraint factor with the current state and historical state evolution trajectory of the human body, generating a set of real-time adjustable constraint boundary parameters. The probability value, weight coefficient, and evolution trend parameters of each failure constraint factor are obtained and integrated with the current multidimensional state vector of the human body, including behavioral indicators, posture deviation values, heart rate, skin conductance, etc., as well as the historical state evolution trajectory. A time-weighted fusion method, such as exponential weighted moving average or Kalman filtering, is used to smooth short-term fluctuations and retain long-term trend information. The probability value of the failure constraint factor is converted into the adjustment range of the corresponding boundary parameter through a linear or nonlinear mapping function. Then, the sensitivity of the boundary parameter is weighted by combining the weight of each factor and the trend parameter to obtain the upper and lower limits of each risk zone and its evolution weight, thereby generating a set of constraint boundary parameters. This set is updated in real time in each sampling period and input into the irradiation adjustment module to dynamically limit the intensity, direction, and rhythm of the physiotherapy lamp, so that the evolution of the human body state is always kept within a safe and controllable range. At the same time, closed-loop correction continues in the next sampling period to achieve real-time adjustable and adaptive control.

[0028] S3: Based on the set of constraint boundary parameters, determine the state feasible region corresponding to the irradiation adjustment strategy, mark the state region that exceeds the state feasible region as the failure risk zone, and construct a three-layer constraint space structure that limits the human body state evolution path.

[0029] The process of determining the feasible region of the illumination adjustment strategy in S3 is as follows: By taking the set of constraint boundary parameters as input and combining them with the current multidimensional state characteristics of the human body, a multidimensional illumination parameter mapping matrix is ​​constructed in the three-layer constraint space. The system acquires historical treatment data and real-time collected multi-dimensional signals such as human behavioral responses, heart rate, skin conductance, and postural stability. These signals are then time-synchronized, outlier-removed, and normalized to form a standardized human state vector. Previously extracted treatment failure constraint factors, such as radiation deviation probability, resistance reaction probability, and radiation sensitivity, are mapped to a set of constraint boundary parameters. The three-layer constraint space includes a low-risk layer, an adjustable-risk layer, and a high-risk layer. The system discretizes parameters such as light intensity, irradiation angle, wavelength, and irradiation time in the three-layer space using a grid partitioning method. Each parameter value is matched with the human state vector to form a four-dimensional irradiation parameter mapping matrix. Each cell contains a parameter value, a corresponding risk layer identifier, and a human state adaptation value, achieving a multi-dimensional and quantifiable parameter distribution representation.

[0030] The parameters of light intensity, wavelength, illumination angle, and duration are mapped to low-risk, adjustable-risk, and high-risk zones. For each parameter dimension in the irradiation parameter matrix, a risk index corresponding to each parameter value is calculated based on historical human irradiation data and failure constraint probability. The light intensity value is quantified according to the photosensitivity response curve. The range below the safety threshold is marked as the low-risk zone, the intermediate range is marked as the adjustable risk zone, and the range above the maximum safety threshold is marked as the high-risk zone. The wavelength is calculated based on the tissue absorption coefficient and skin penetration depth curve to determine the corresponding biological effect intensity, and then mapped to the three-layer risk zone. The irradiation angle is combined with the posture offset value and the local light-sensitive area for projection to calculate the degree of irradiation deviation and map it to the risk zone. The duration is determined based on the cumulative value of irradiation response and the heat accumulation model to determine the risk of excessive irradiation, and is mapped to the three-layer risk zone accordingly. The risk zone mapping results of each parameter form a multi-dimensional risk matrix, providing basic data for subsequent fusion and optimization.

[0031] Weighted fusion calculations are performed on each parameter dimension and its corresponding risk zone, incorporating the probability values, evolution trends, and human state sensitivity of the constraint boundary parameters into the weighting factors. For each parameter dimension, initial weight coefficients are assigned to low-risk, adjustable-risk, and high-risk zones. These coefficients are then dynamically adjusted based on the failure probability of the constraint boundary parameters, historical evolution trends, and current human body state sensitivity. The probability values ​​are obtained through statistical analysis of historical treatment data, and the evolution trends are calculated using exponentially weighted moving averages or Kalman filters to reflect the short-term and long-term impacts of parameter changes on failure risk. Human body state sensitivity is calculated using normalized heart rate, skin conductance response, and posture deviation values ​​to represent the weight of different parameters on the immediate human response. Finally, the risk indices and weight factors of each dimension are weighted and summed to form a multi-dimensional weighted fusion matrix, achieving a comprehensive quantitative representation of parameter values, risk levels, and human body state.

[0032] The upper and lower limits, allowable change step size and safety margin of each parameter are calculated by calling a multi-objective optimization algorithm to form a structured state feasible region; Based on the weighted fusion matrix, a multi-objective optimization algorithm is used to solve for the optimal parameter range. The optimization objectives include satisfying the constraint boundary conditions, minimizing risk, and ensuring parameter continuity. During the solution process, particle swarm optimization or multi-objective genetic algorithm is used, combined with human state vector constraints to filter the solution space. The optimization results output the allowable upper and lower limits, recommended change step size, and safety margin value for each parameter. The safety margin is calculated based on historical treatment data and physiological response curves to ensure that the feasible region under each parameter combination will not touch the high-risk area. Finally, the optimized range of light intensity, wavelength, irradiation angle, and irradiation time is structured and stored to form a structured feasible region, which can be called in real time by the irradiation adjustment and control module to achieve fine parameter adjustment.

[0033] The process of constructing a three-layer constraint space structure that limits the evolution path of human state in S3 is as follows: The combined illumination parameters outside the feasible state domain are calibrated as the failure risk zone, and low-risk, adjustable-risk, and high-risk parameter levels are established in the three-layer constraint space. Obtain the state feasible region matrix, which includes the allowable range and optimization step size of multi-dimensional parameters such as light intensity, wavelength, illumination angle, and illumination time. Exhaustively enumerate or stratify all combined parameters to determine which combinations exceed the upper and lower limits of the state feasible region and mark them as failure risk zones. Classify parameter combinations according to risk probability: the low-risk layer contains parameters with probability values ​​lower than the historical statistical threshold, the middle layer is the adjustable risk layer, and the high-risk layer corresponds to parameter combinations with a high failure probability. The risk level is stored in the form of a three-dimensional matrix, with each cell containing parameter combination value, risk label, and historical failure probability, providing basic data for state evolution path constraints. Referring to phototherapy experimental data, combinations with light intensity below 200 milliwatts per square centimeter are marked as low risk, 200 to 350 milliwatts per square centimeter as adjustable risk, and above 350 milliwatts per square centimeter as high risk. Map other parameters proportionally to the corresponding risk layers.

[0034] To address the multidimensional state characteristics of the human body, each dimension is mapped to a three-layer constraint space, and boundary functions, variation ranges, and evolution weights are defined for each layer. By collecting real-time multidimensional human state signals, including posture deviation values, heart rate, skin conductance, and behavioral response indicators, each state dimension is normalized and standardized, and then mapped to the parameter dimensions corresponding to the three-layer constraint space. Boundary functions are defined for each layer, and the progression relationship of parameters from the low-risk layer to the adjustable-risk layer and the high-risk layer is described by linear or nonlinear functions. The variation range represents the upper and lower limits of the parameter's safe fluctuation within the layer, and the evolution weight represents the relative influence of the parameter on path selection during state evolution. The weight calculation can be based on sensitivity analysis of historical treatment data.

[0035] Define legal transition rules for state evolution paths in the constraint space, including the allowed step size for increasing or decreasing state dimensions, the direction of parameter combination changes, and cross-layer transition conditions; Within the three-layer constraint space, an evolution step size is set for each parameter dimension, which is the allowable increase or decrease of the parameter in a continuous time step. This magnitude is obtained based on historical irradiation data statistics and can be determined by calculating the mean and standard deviation through a sliding window. The direction of parameter combination change is determined by multi-dimensional gradient calculation to determine the safe adjustment direction of each dimension, ensuring that the path migration always proceeds along the low-risk or controllable-risk direction. Cross-layer migration conditions are controlled by probability thresholds and cumulative risk limits.

[0036] Dynamic updates are performed on boundary parameters and migration rules, and weighted corrections are made based on real-time multi-source signals to form a complete three-layer constraint space structure. During treatment, multiple signals such as heart rate, skin conductance, and posture stability are collected in real time. The signals are processed by exponential weighted moving average or Kalman filter to reduce the impact of short-term fluctuations. Based on these signals, the boundary functions and evolution weights of each parameter dimension are dynamically corrected, the upper and lower limits and migration step size of each layer parameter are updated, and the cross-layer migration conditions are adjusted to adapt to the human body's immediate response and achieve continuous constraint and dynamic control of the human body's state evolution path.

[0037] S4: Based on a three-layer constrained spatial structure, continuously track the human body's state evolution trajectory, and when the trajectory deviates from the failure risk zone or the human body shows an abnormal trend in subjective feelings, adjust the weights of physiotherapy orientation and comfort orientation and execute comfort orientation adjustment to generate joint adjustment instructions.

[0038] The process of continuously tracking the human body's state evolution trajectory based on a three-layer constrained spatial structure in S4 is as follows: Collect and record multimodal response data of the human body during irradiation, including behavioral response, physiological reaction and posture stability signals, and map each state dimension to a three-layer constraint space structure; During irradiation, human behavioral response signals, physiological reaction signals, and posture stability signals are collected in parallel by multi-channel sensors. The behavioral response signals include the amplitude and direction of limb movement, the physiological reaction signals include heart rate, skin conductance, and blood oxygen concentration, and the posture stability signals are collected by inertial measurement units or optical tracking. The collected signals are synchronously corrected in the time dimension, and outliers are filtered by median filtering or three times the standard deviation. Each dimension is standardized to normalize the signals to a unified dimension. The normalized value of each state dimension is projected onto a three-layer constraint space structure through a mapping function to determine its location in the low-risk layer, adjustable-risk layer, or high-risk layer, providing basic spatial coordinates for real-time trajectory analysis. For example, heart rate changes in the range of 50 to 80 beats per minute are mapped to the low-risk layer, 80 to 100 beats per minute to the adjustable-risk layer, and above 100 beats per minute to the high-risk layer.

[0039] Calculate the numerical position, deviation, and cross-layer migration risk index of each state dimension in the constraint space to generate a multi-dimensional state deviation matrix; For each state dimension, its relative position value in the three-layer constraint space is compared with the boundary function of each layer to calculate the deviation, which is the percentage distance between the current state and the boundary of the low-risk layer. At the same time, the potential risk of cross-layer migration is assessed. The cross-layer migration risk index is calculated by combining historical failure probability data and current human sensitivity data to reflect the possibility of parameters migrating from a low-risk or adjustable-risk layer to a high-risk layer. The parameters of light intensity, wavelength, irradiation angle and irradiation time are coupled with the multi-dimensional state dimensions of the human body to form a multi-dimensional state deviation matrix. Each matrix unit contains state value, deviation, cross-layer risk and timestamp to realize a quantitative description of state deviation. The mean and standard deviation of attitude offsets sampled continuously within five minutes can be statistically analyzed. The deviation of each sampling point is calculated by formula and combined with the cross-layer migration probability table to form a matrix.

[0040] Based on the legal migration rules of the evolution path defined in the three-layer constraint space, and combined with the multi-dimensional state deviation matrix, the human state trajectory is gradually tracked and analyzed in time to identify the trajectory sequence of potential deviation failure risk areas. At fixed time intervals, each collected human state point is mapped to a three-layer constraint space. Based on the deviation matrix and boundary function of each state point, its legality within the current risk layer is determined. Combined with the legal migration rules of the evolution path, including the allowed step size for increasing or decreasing the state dimension, the direction of parameter combination change, and cross-layer migration conditions, a trajectory sequence is formed for continuous state points. In the trajectory sequence, when some state points deviate from the low-risk layer and continue to approach the boundary of the adjustable or high-risk layer, they are marked as potential failure risk zone trajectory sequences. The continuity of the trajectory, deviation trend, and cross-layer migration probability can be calculated by sliding window or recursive algorithm to ensure quantitative tracking of potential risk paths.

[0041] Data annotation and indexing are performed on the trajectory sequences to generate a trajectory mapping table; Each trajectory sequence is labeled according to time series and risk level, recording the trajectory start time, end time, involved parameter dimensions, deviation, and cross-level migration risk indicators. At the same time, a unique index number is generated to form a searchable trajectory mapping table. The trajectory mapping table can be used for subsequent analysis, model training, or dynamic adjustment strategy input to quickly locate potential risk trajectories and their evolution patterns. In actual data management, the trajectory table can be stored as a multidimensional array or relational database. Each record contains four-dimensional parameter status, deviation matrix value, risk level identifier, and timestamp, supporting real-time query and historical trajectory statistical analysis.

[0042] The process of generating joint control commands in S4 is as follows: Using the multidimensional state deviation matrix and trajectory mapping table as input, the deviation magnitude, direction and cross-layer migration risk of each state dimension in the three-layer constraint space are quantified. Using a multidimensional state deviation matrix and trajectory mapping table as input, the deviation magnitude of each state dimension in the three-layer constraint space is calculated as the percentage difference between the current state value and the boundary value of the low-risk layer or adjustable-risk layer. The deviation direction is obtained by calculating the vector direction of the state value relative to the target feasible region, reflecting whether the current parameter should move towards the lower-risk or adjustable-risk layer. The cross-layer migration risk is obtained through historical trajectory statistics, including the probability distribution of state points moving from the low-risk layer to the adjustable-risk or high-risk layer, as well as the sensitivity coefficient of each state dimension to cross-layer migration. Finally, a quantitative table is formed, with each record containing the state dimension identifier, deviation magnitude, deviation direction, cross-layer migration probability, and timestamp, realizing a quantitative description of human state deviation and potential risks.

[0043] The quantification results are weighted and combined with the historical evolution information of the trajectory mapping table to construct a deviation evaluation characterization; Each quantified record is weighted, with weighting factors including state dimension sensitivity, duration and frequency of deviations in historical trajectories, cross-layer migration risk index, and the importance of each state dimension in joint regulation. Historical evolution information is extracted through a trajectory mapping table. For example, the number of times a certain state dimension deviates from the low-risk layer continuously in past irradiation cycles and the cumulative magnitude are counted and converted into weighting factors. The weighted deviation values, combined with the deviation direction and cross-layer risk, generate a deviation assessment representation matrix. Each state dimension corresponds to a comprehensive deviation index and correction direction, providing calculable input for the joint regulation algorithm. In the case study, when the heart rate deviates from the low-risk layer for more than five consecutive samplings and the light intensity deviation is large, the weighted deviation assessment index can reach 70%, providing a quantitative basis for adjustment instructions.

[0044] Based on the deviation assessment characterization, the irradiation guidance control parameters and the weights of each state dimension are calculated according to the preset weight adjustment algorithm, and a joint adjustment command is generated. Using deviation assessment as input, a pre-defined weighted adjustment algorithm is employed to convert the comprehensive deviation index of each state dimension into specific illumination guidance control parameter adjustment values, including light intensity adjustment amplitude, wavelength correction value, illumination angle adjustment step size, and illumination time correction amount. Simultaneously, the algorithm generates the weight allocation of each state dimension in the joint adjustment to ensure that parameter adjustments are coordinated and consistent across dimensions. The algorithm calculates instructions through linear weighting or nonlinear mapping to ensure joint adjustment is achieved under the safety conditions of the three-layer constraint space. After the joint adjustment instructions are formed, they are sent to the illumination control module in a time series to achieve simultaneous adjustment of multi-dimensional parameters.

[0045] S5: Based on the trajectory corrected by the joint adjustment command, combined with the determined state feasible region, the irradiation control mode is dynamically switched, and the failure constraint factor and state feasible region parameters are updated synchronously during the mode switching process to form the updated irradiation control state.

[0046] The process of forming the updated irradiation control state in S5 is as follows: Based on the joint adjustment instructions and the current status, determine the applicable irradiation mode and corresponding parameter combination; The system receives joint adjustment commands, including light intensity adjustment range, irradiation angle correction, wavelength selection, and irradiation time adjustment. Simultaneously, it acquires the current human body state vector, including posture deviation, heart rate, skin conductance, and behavioral response signals. By matching predefined mode parameter combinations in the irradiation mode library, it calculates the fit degree between each mode and the current human body state. The fit degree calculation considers the risk layer position of the parameters in the three-layer constraint space, the deviation index of each dimension, and the weight coefficients assigned in the joint adjustment commands. The optimal irradiation mode is selected using a weighted similarity scoring method. The finally selected mode corresponds to a specific set of irradiation parameter combinations, ensuring that each parameter is in a low-risk or controllable risk zone under the current state, while also meeting the deviation correction requirements of the joint adjustment commands.

[0047] The failure constraint factor and state feasible domain parameters are calculated and updated in real time, and the current human state is mapped to the feasible space of the selected mode. After the irradiation mode is determined, failure constraint factors are calculated in real time for each parameter dimension, including the probability of irradiation deviation, the probability of resistance reaction, and the cumulative risk index. The calculation method combines historical treatment data with currently collected physiological, behavioral, and postural signals. Short-term and long-term trends are obtained through sliding window statistics and exponentially weighted moving average. Based on the updated constraint factors, the feasible domain parameters are recalculated, including the upper and lower limits, allowable step size, and safety margin for each parameter. The current human state vector is mapped to the feasible space of the selected irradiation mode, forming a correspondence matrix between the current parameter state and the feasible domain.

[0048] Record the state mapping, irradiation parameter execution and constraint parameter update during the mode switching process, and output the updated irradiation control state, including irradiation mode, parameter configuration and state feasible domain mapping information; During mode switching and parameter adjustment, each parameter update is recorded, including the current human body state mapping, the executed irradiation parameter values, the failure constraint factor update values, and the real-time changes in the state feasible domain. The recorded content is stored in time series and can be used for trajectory analysis, model training, or adjustment strategy optimization. The final updated irradiation control state consists of three parts: the selected irradiation mode identifier, the specific parameter configuration values, and the mapping matrix of the current state in the mode feasible space. This data structure enables real-time monitoring and tracking, ensuring the safety and continuity of parameter adjustment during continuous irradiation.

[0049] Example 2: Please refer to Figure 2 As shown, a physiotherapy lamp irradiation adjustment system based on human body state perception includes: State modeling module: Parallel acquisition of signals such as human behavior and heart rate, synchronization, anomaly removal and normalization, and construction of human irradiation response evolution characteristic model; Constraint Extraction Module: Based on historical physiotherapy data and the current state, it extracts failure constraint factors such as failure risk, deviation trend, and resistance probability from the physiotherapy state model and maps them into a set of constraint boundary parameters; Feasible domain delineation module: Based on the set of constraint boundary parameters, determine the state feasible domain of the illumination strategy, mark the region outside the feasible domain as the failure risk zone, and construct a three-layer constraint space to limit the state evolution path; Trajectory Analysis Module: Continuously tracks the evolution trajectory of the human body in the three-layer constraint space, and adjusts the irradiation weight to generate joint adjustment instructions when the body deviates from the failure risk zone or an abnormal trend occurs; Mode Management Module: Based on the human body state trajectory adjusted by the joint adjustment command, switch the irradiation mode and update the failure constraint factor and state feasible domain, and generate the updated irradiation control state.

[0050] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0051] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for adjusting the irradiation of a physiotherapy lamp based on human body state perception, characterized in that, Includes the following steps: Parallel acquisition of behavioral response signals, physiological reaction signals, and posture stability signals of the irradiated human body; time synchronization, anomaly removal, and scale normalization of the signals; and construction of a physiotherapy state model characterizing the evolution of the human body's irradiation response. Based on historical physiotherapy data and current human body condition, failure constraint factors reflecting the risk of physiotherapy failure, the trend of radiation deviation, and the probability of human body resistance reaction are extracted from the physiotherapy state model and mapped to the set of constraint boundary parameters in the physiotherapy state model. Based on the set of constraint boundary parameters, the state feasible region corresponding to the irradiation adjustment strategy is determined, and the state region that exceeds the state feasible region is marked as the failure risk zone, thus constructing a three-layer constraint space structure that limits the human state evolution path. Based on a three-layer constrained spatial structure, the system continuously tracks the trajectory of human body state evolution. When the trajectory deviates from the failure risk zone or when the human body shows an abnormal trend in subjective feelings, the system adjusts the weights of physiotherapy orientation and comfort orientation and executes comfort orientation adjustment to generate joint adjustment instructions. Based on the trajectory corrected by the joint adjustment command, and combined with the determined state feasible region, the irradiation control mode is dynamically switched, and the failure constraint factor and state feasible region parameters are updated synchronously during the mode switching process to form the updated irradiation control state.

2. The method for adjusting the irradiation of a physiotherapy lamp based on human body state perception according to claim 1, characterized in that, The process of constructing a physical therapy state model that characterizes the evolution of the human body's response to radiation is as follows: Deploy multi-channel sensor acquisition nodes within the irradiation area, including motion sensors, heart rate and skin conductance sensors, posture detection devices, and ambient light intensity sensors; The signals from each sensor are synchronized with microsecond-level timestamps, and signal consistency is ensured through filtering, outlier removal, and normalization algorithms. Based on the processed multi-source signals, a multi-dimensional temporal feature extraction and weighted fusion method is used to generate a feature sequence that reflects the evolution of the human body's response to irradiation, and a physiotherapy state model that characterizes the evolution of the human body's response to irradiation is constructed accordingly.

3. The method for adjusting the irradiation of a physiotherapy lamp based on human body state perception according to claim 2, characterized in that, The process of extracting failure constraint factors reflecting the risk of physiotherapy failure, the trend of radiation deviation, and the probability of human resistance from the physiotherapy state model is as follows: Input historical data and the current human body status into the constructed physiotherapy status model; By using statistical modeling and Bayesian update methods, the probability distribution and trend estimate of human state deviation under different irradiation conditions are generated. The probability distribution and trend are used to construct a multidimensional state representation of the physiotherapy state model. As input, multi-task regression analysis and reinforcement learning algorithms are applied to predict potential resistance responses and generate quantitative failure constraint factors.

4. The method for adjusting the irradiation of a physiotherapy lamp based on human body state perception according to claim 3, characterized in that, The process of mapping to the set of constraint boundary parameters in the physiotherapy state model is as follows: Based on the multidimensional state characteristics of the human body, the failure constraint factors are divided into three layers of constraint space according to state sensitivity and time dependence, including low risk, adjustable risk and high risk areas. Define specific numerical ranges and evolution weights for each boundary layer; A dynamic mapping algorithm is executed for each boundary parameter to weight and fuse the probability value, weight, and trend parameter of the failure constraint factor with the current state and historical state evolution trajectory of the human body, generating a set of real-time adjustable constraint boundary parameters.

5. The method for adjusting the irradiation of a physiotherapy lamp based on human body state perception according to claim 4, characterized in that, The process of determining the feasible region of the illumination adjustment strategy is as follows: By taking the set of constraint boundary parameters as input and combining them with the current multidimensional state characteristics of the human body, a multidimensional illumination parameter mapping matrix is ​​constructed in the three-layer constraint space. The parameters of light intensity, wavelength, illumination angle, and duration are mapped to low-risk, adjustable-risk, and high-risk zones. Weighted fusion calculations are performed on each parameter dimension and its corresponding risk zone, incorporating the probability values, evolution trends, and human state sensitivity of the constraint boundary parameters into the weighting factors. The upper and lower limits, allowable step size, and safety margin of each parameter are calculated by calling a multi-objective optimization algorithm to form a structured feasible state region.

6. The method for adjusting the irradiation of a physiotherapy lamp based on human body state perception according to claim 5, characterized in that, The process of constructing a three-layer constrained spatial structure that limits the evolutionary path of the human body is as follows: The combined illumination parameters outside the feasible state domain are calibrated as the failure risk zone, and low-risk, adjustable-risk, and high-risk parameter levels are established in the three-layer constraint space. To address the multidimensional state characteristics of the human body, each dimension is mapped to a three-layer constraint space, and boundary functions, variation ranges, and evolution weights are defined for each layer. Define legal transition rules for state evolution paths in the constraint space, including the allowed step size for increasing or decreasing state dimensions, the direction of parameter combination changes, and cross-layer transition conditions; The boundary parameters and migration rules are dynamically updated, and weighted corrections are performed based on real-time multi-source signals to form a complete three-layer constraint space structure.

7. The method for adjusting the irradiation of a physiotherapy lamp based on human body state perception according to claim 6, characterized in that, The process of continuously tracking the human body's state evolution trajectory based on a three-layer constrained spatial structure is as follows: Collect and record multimodal response data of the human body during irradiation, including behavioral response, physiological reaction and posture stability signals, and map each state dimension to a three-layer constraint space structure; Calculate the numerical position, deviation, and cross-layer migration risk index of each state dimension in the constraint space to generate a multi-dimensional state deviation matrix; Based on the legal migration rules of the evolution path defined in the three-layer constraint space, and combined with the multi-dimensional state deviation matrix, the human state trajectory is gradually tracked and analyzed in time to identify the trajectory sequence of potential deviation failure risk areas. Data annotation and indexing are performed on the trajectory sequence to generate a trajectory mapping table.

8. The method for adjusting the irradiation of a physiotherapy lamp based on human body state perception according to claim 7, characterized in that, The process of generating joint control commands is as follows: Using the multidimensional state deviation matrix and trajectory mapping table as input, the deviation magnitude, direction and cross-layer migration risk of each state dimension in the three-layer constraint space are quantified. The quantification results are weighted and combined with the historical evolution information of the trajectory mapping table to construct a deviation evaluation characterization; Based on the deviation assessment characterization, the irradiation guidance control parameters and the weights of each state dimension are calculated according to the preset weight adjustment algorithm, and a joint adjustment command is generated.

9. The method for adjusting the irradiation of a physiotherapy lamp based on human body state perception according to claim 8, characterized in that, The process of forming the updated irradiation control state is as follows: Based on the joint adjustment instructions and the current status, determine the applicable irradiation mode and corresponding parameter combination; The failure constraint factor and state feasible domain parameters are calculated and updated in real time, and the current human state is mapped to the feasible space of the selected mode. Record the state mapping, irradiation parameter execution, and constraint parameter update during the mode switching process, and output the updated irradiation control state, including irradiation mode, parameter configuration, and state feasible domain mapping information.

10. A physiotherapy lamp irradiation adjustment system based on human body state perception, applied to the method described in any one of claims 1-9, characterized in that, include: State modeling module: Parallel acquisition of signals such as human behavior and heart rate, synchronization, anomaly removal and normalization, and construction of human irradiation response evolution characteristic model; Constraint Extraction Module: Based on historical physiotherapy data and the current state, it extracts failure constraint factors such as failure risk, deviation trend, and resistance probability from the physiotherapy state model and maps them into a set of constraint boundary parameters; Feasible domain delineation module: Based on the set of constraint boundary parameters, determine the state feasible domain of the illumination strategy, mark the region outside the feasible domain as the failure risk zone, and construct a three-layer constraint space to limit the state evolution path; Trajectory Analysis Module: Continuously tracks the evolution trajectory of the human body in the three-layer constraint space, and adjusts the irradiation weight to generate joint adjustment instructions when the body deviates from the failure risk zone or an abnormal trend occurs; Mode Management Module: Based on the human body state trajectory adjusted by the joint adjustment command, switch the irradiation mode and update the failure constraint factor and state feasible domain, and generate the updated irradiation control state.

Images