Rehabilitation training method, device, equipment, medium and program product

Abstract

This invention relates to a rehabilitation training method, apparatus, equipment, medium, and program product. The method includes acquiring initial state data of a target subject, obtaining impedance configuration parameters from a rehabilitation training institution based on the initial state data, and controlling the rehabilitation training institution to perform passive limb training on the target subject based on the impedance configuration parameters. During the passive training phase, real-time state data and real-time torque data of the target subject are collected, and an active force characterization value is calculated based on the real-time torque data. A bilateral movement deviation value is also calculated based on the real-time state data. The real-time state data collected during the passive training phase is used to assess the degree of motor completion. When the assessment result reaches the stage transition standard, the rehabilitation training institution is switched to an active assisted training mode to continue training until a training end command is received. This invention enables the matching of training resistance and movement trajectory with the patient's real-time force exertion ability and motor coordination.

Description

Technical Field

[0001] This invention relates to the technical field of rehabilitation training, and in particular to a rehabilitation training method, device, equipment, medium, and program product. Background Technology

[0002] In the field of limb-mobility rehabilitation training, existing technologies generally suffer from insufficient dynamic adaptation capabilities. On the one hand, training resistance is often adjusted manually based on fixed levels or the therapist's experience, failing to dynamically adjust according to real-time physiological characteristics such as the patient's muscle exertion and joint range of motion. This leads to a disconnect between training intensity and the patient's actual ability, easily resulting in undertraining or secondary injury. On the other hand, limb movement trajectories are preset to fixed paths. When the patient's motor ability in a certain limb declines, the system continues to drive along the original trajectory, causing overcompensation on the healthy side and poor training effect on the affected side. Furthermore, the lack of in-depth analysis of continuous movement and physiological data makes it difficult for therapists to accurately quantify rehabilitation progress. Summary of the Invention

[0003] The main objective of this invention is to provide a rehabilitation training method, device, equipment, medium, and program product. By constructing a dynamic closed-loop adaptation mechanism between training parameters and the patient's real-time state, the invention achieves millisecond-level matching between training resistance and movement trajectory and the patient's real-time force exertion ability and motor coordination, thereby enabling dynamic adjustment of training parameters according to the patient's real-time state.

[0004] To achieve the above objectives, the present invention provides a rehabilitation training method, comprising: The initial state data of the target object is obtained, and impedance configuration parameters are obtained from the rehabilitation training institution based on the initial state data. Based on the impedance configuration parameters, the rehabilitation training institution is controlled to perform passive limb training on the target object. During the passive training phase, real-time state data and real-time torque data of the target object are collected, and active force characterization value is calculated based on the real-time torque data, and bilateral motion deviation value is calculated based on the real-time state data. Based on the active force characterization value and the bilateral motion deviation value, the resistance and trajectory of the rehabilitation training mechanism are continuously dynamically adjusted during the passive training phase; The real-time status data collected during the passive training phase is used to assess the degree of exercise completion. When the assessment result reaches the phase transition standard, the rehabilitation training institution is switched to active assisted training mode to continue training until a training end instruction is received.

[0005] Further, the step of acquiring initial state data of the target object and obtaining impedance configuration parameters from the rehabilitation training institution based on the initial state data includes: The rehabilitation training institution collects the initial joint angle values ​​of the target object's limbs, as well as the initial surface electromyography (EMG) signals output by the surface EMG sensors attached to the preset force-generating muscle groups of the target object's limbs. The initial surface electromyography (EMG) signals are filtered and feature extracted to obtain initial EMG feature parameters. The initial joint angle values ​​of each limb are integrated with the initial electromyographic characteristic parameters to obtain the initial state data; Based on the initial state data, the rehabilitation stage in the rehabilitation training institution is matched to obtain the rehabilitation stage to which the target object belongs, and the impedance configuration parameters of the rehabilitation stage are extracted.

[0006] Furthermore, in the passive training phase, real-time state data and real-time torque data of the target object are collected, and an active force characterization value is calculated based on the real-time torque data. The bilateral motion deviation value is also calculated based on the real-time state data, including: The real-time status data collected by the rehabilitation training institution during the passive training phase is obtained, and the motion trajectory parameters of the first limb and the second limb are extracted from the real-time status data. For each symmetrical limb pair, the trajectory deviation value between the motion trajectory parameters of the first limb and the second limb at the same acquisition time is calculated, and the trajectory deviation values ​​at multiple consecutive acquisition times are averaged to obtain the bilateral motion deviation value of each symmetrical limb pair. The real-time torque data output by the drive motors of each joint in the rehabilitation training institution is obtained, and the difference between the real-time torque value at each acquisition time and the preset passive training reference torque value is calculated to obtain the torque deviation value at each acquisition time. The torque deviation values ​​at multiple consecutive acquisition times are accumulated to obtain the active force characterization value.

[0007] Furthermore, the step of continuously and dynamically adjusting the resistance and trajectory of the rehabilitation training mechanism during the passive training phase based on the active force representation value and the bilateral motion deviation value includes: Each of the active force exertion characterization values ​​is compared with a preset desired force exertion range. When the active force exertion characterization value deviates from the desired force exertion range, the corresponding deviation state is obtained, and a resistance adjustment control signal is generated based on the deviation state. The resistance adjustment control signal is output to the resistance execution unit of the rehabilitation training institution to perform resistance adjustment; When any of the bilateral motion deviation values ​​is greater than a preset deviation tolerance threshold, the degree of deviation between the real-time status data corresponding to the bilateral motion deviation value and the preset joint trajectory is identified, and the deviation-dominant side and the deviation-following side are determined according to the degree of deviation. The trajectory compensation amount of the deviation-following side is calculated based on the deviation-dominant side and the preset desired angle value, and the motion trajectory of the deviation-following side is corrected based on the trajectory compensation amount.

[0008] Further, when any of the bilateral motion deviation values ​​exceeds a preset deviation tolerance threshold, identifying the degree of deviation between the real-time state data corresponding to the bilateral motion deviation value and the preset joint trajectory, and determining the deviation-dominant side and the deviation-following side based on the degree of deviation, includes: When the bilateral motion deviation value is greater than the deviation tolerance threshold, locate the real-time status data corresponding to the bilateral motion deviation value, and extract the joint angle value of the first limb and the joint angle value of the second limb from the real-time status data. Calculate the absolute value of the first angle difference between the joint angle value of the first limb and the first expected angle value at the corresponding moment in the preset joint trajectory; calculate the absolute value of the second angle difference between the joint angle value of the second limb and the second expected angle value at the corresponding moment in the preset joint trajectory. When the absolute value of the first angle difference is greater than the absolute value of the second angle difference, the first limb is determined to be the deviation-dominant side, and the second limb is determined to be the deviation-following side. Conversely, the second limb is determined as the deviation-dominant side, and the first limb is determined as the deviation-following side.

[0009] Furthermore, the real-time status data collected during the passive training phase is used to assess the degree of motor completion. When the assessment result reaches the phase transition standard, the rehabilitation training institution is switched to active assisted training mode to continue training until a training end instruction is received, including: The motion trajectory data of multiple consecutive acquisition times are extracted from the real-time status data, and each of the motion trajectory data is compared with the preset expected motion trajectory to identify the motion matching deviation at each acquisition time. When the motion matching deviation is less than the preset deviation threshold at multiple consecutive acquisition times, the evaluation result is determined to have reached the stage transition standard. A phase switching instruction is generated based on the phase transition standard, and an impedance adjustment factor matching the current real-time status data is extracted from the rehabilitation training institution. Based on the impedance adjustment factor, the impedance configuration parameter is adjusted to an initial value that adapts to the active assisted training mode. In response to the stage switching command, the rehabilitation training institution is controlled to switch to the active assisted training mode with the adjusted impedance configuration parameters, and the training continues until a training end command is received.

[0010] The present invention also provides a rehabilitation training device, applied to any one of the rehabilitation training methods described above, comprising: The acquisition module is used to acquire initial state data of the target object, and obtain impedance configuration parameters from the rehabilitation training institution based on the initial state data, and control the rehabilitation training institution to perform passive limb training on the target object based on the impedance configuration parameters. The analysis module is used to collect real-time state data and real-time torque data of the target object during the passive training phase, calculate the active force characterization value based on the real-time torque data, and calculate the bilateral motion deviation value based on the real-time state data. The association module is used to continuously and dynamically adjust the resistance and trajectory of the rehabilitation training mechanism during the passive training phase based on the active force characterization value and the bilateral motion deviation value. The processing module is used to evaluate the exercise completion rate of the real-time status data collected during the passive training phase. When the evaluation result reaches the phase transition standard, the rehabilitation training institution is switched to the active assisted training mode to continue training until the training end instruction is received.

[0011] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the rehabilitation training method as described in any of the preceding claims.

[0012] The present invention also provides a readable storage medium storing a computer program that, when executed by a processor, implements the steps of the rehabilitation training method as described in any of the preceding claims.

[0013] The present invention also provides a computer program product comprising a computer program that, when executed by a processor, enables the implementation of the steps of the rehabilitation training method as described in any of the preceding claims.

[0014] The rehabilitation training method, device, equipment, medium, and program product provided by this invention have the following beneficial effects: By synchronously collecting real-time status and torque data of the target object, and calculating the active force representation value based on the real-time torque data, dynamic matching between training resistance and the patient's real-time force exertion ability is achieved. This solves the problem of training intensity being out of sync with the patient's actual ability due to lag in resistance adjustment in existing technologies, and avoids insufficient training or secondary injury. By calculating bilateral movement deviation values ​​based on real-time status data, and continuously adjusting the trajectory of the rehabilitation training institution based on these deviation values ​​during the passive training phase, adaptive calibration of the limb coordinated movement trajectory is achieved. This effectively suppresses excessive compensation behavior of the healthy limbs and ensures that the affected limbs receive sufficient training load. By evaluating the movement completion rate of the real-time status data collected during the passive training phase, and switching to active assisted training mode when the evaluation results reach the phase transition standard, continuous and fine-grained data collection and evaluation of the training process are achieved. This provides therapists with traceable and quantifiable evidence of rehabilitation progress, supporting the refined adjustment of training programs. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of a rehabilitation training system according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating a rehabilitation training method according to an embodiment of the present invention; Figure 3 This is a flowchart illustrating a rehabilitation training method according to an embodiment of the present invention; Figure 4 This is a flowchart illustrating a rehabilitation training method according to an embodiment of the present invention; Figure 5 This is a flowchart illustrating a rehabilitation training method according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the structure of a rehabilitation training device in one embodiment of the present invention; Figure 7 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention.

[0017] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] 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, not all, of the embodiments of the present invention. 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.

[0019] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof. It should also be understood that, as used in this specification and the appended claims, the term "and / or" refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0020] Furthermore, in the description of this invention and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0021] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of the invention include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0022] It should be understood that the sequence number of each step in the following embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0023] To illustrate the technical solution of the present invention, specific embodiments are described below.

[0024] In the field of limb-mobility rehabilitation training, existing technologies generally suffer from insufficient dynamic adaptation capabilities. On the one hand, training resistance is often adjusted manually based on fixed levels or the therapist's experience, failing to dynamically adjust according to real-time physiological characteristics such as the patient's muscle exertion and joint range of motion. This leads to a disconnect between training intensity and the patient's actual ability, easily resulting in undertraining or secondary injury. On the other hand, limb movement trajectories are preset to fixed paths. When the patient's motor ability in a certain limb declines, the system continues to drive along the original trajectory, causing overcompensation on the healthy side and poor training effect on the affected side. Furthermore, the lack of in-depth analysis of continuous movement and physiological data makes it difficult for therapists to accurately quantify rehabilitation progress.

[0025] To address the aforementioned issues, this application proposes a rehabilitation training method, device, equipment, medium, and program product. By synchronously collecting real-time status data and real-time torque data of the target object, and calculating the active force expression value based on the real-time torque data, dynamic matching of training resistance and the patient's real-time force expression ability is achieved. This solves the problem of training intensity being out of sync with the patient's actual ability due to lag in resistance adjustment in existing technologies, thus avoiding insufficient training or secondary injury. By calculating bilateral movement deviation values ​​based on real-time status data, and continuously dynamically adjusting the trajectory of the rehabilitation training mechanism based on these deviation values ​​during the passive training phase, adaptive calibration of the limb coordinated movement trajectory is achieved, effectively suppressing excessive compensatory behavior of the healthy limb and ensuring that the affected limb receives sufficient training load. By evaluating the movement completion rate of the real-time status data collected during the passive training phase, and switching to active assisted training mode when the evaluation result reaches the stage transition standard, continuous and fine-grained data collection and evaluation of the training process are achieved. This provides therapists with traceable and quantifiable evidence of rehabilitation progress, supporting the refined adjustment of training programs.

[0026] The rehabilitation training method provided in this embodiment of the invention can be applied to, for example... Figure 1 The rehabilitation training system shown includes terminal equipment and rehabilitation training devices, wherein the terminal equipment communicates with the server via a network or bus.

[0027] The rehabilitation training device can be a terminal device, which refers to a device that corresponds to a server and provides local services to customers. This terminal device includes, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices.

[0028] The rehabilitation training device can also be a server, which can be implemented using a standalone server or a server cluster composed of multiple servers.

[0029] Reference Figure 2 As shown, the present invention provides a rehabilitation training method, comprising: Step S1: Obtain the initial state data of the target object, and obtain the impedance configuration parameters from the rehabilitation training institution based on the initial state data, and control the rehabilitation training institution to perform passive limb training on the target object based on the impedance configuration parameters; Among them, the rehabilitation training institution refers to the mechanical execution system used to drive the limbs of the target object to perform rehabilitation training movements. It includes a drive motor, transmission mechanism, angle sensors of each joint, surface electromyography sensor, torque sensor and resistance execution unit. The institution controls the limbs to move along the preset desired trajectory according to the drive command.

[0030] Specifically, at the start of training, initial joint angle values ​​of the target subject's limbs are collected using joint angle sensors provided by the rehabilitation training institution. Simultaneously, initial surface electromyography (EMG) signals are collected using surface EMG sensors attached to preset muscle groups on the target subject's limbs. The collected initial EMG signals undergo bandpass filtering to remove noise components outside the 20Hz to 450Hz frequency band. The root mean square value of the filtered signal in the time domain is calculated as the amplitude feature. A Fourier transform is performed on the filtered signal to obtain the power spectrum. The frequency value corresponding to 50% cumulative energy in this power spectrum is calculated as the frequency feature. The amplitude feature and frequency feature are used as EMG characteristic parameters.

[0031] The initial joint angle values ​​and electromyographic characteristic parameters of each limb are combined according to limb channels to construct an initial state vector containing all limb channels. The rehabilitation model library in the rehabilitation training institution is traversed. This library stores impedance configuration parameters corresponding to different rehabilitation stages and their associated standard state vectors. The Euclidean distance between the initial state vector and each standard state vector is calculated, and the impedance configuration parameter corresponding to the standard state vector with the smallest Euclidean distance is selected as the initial impedance configuration parameter. Drive control commands are generated based on this initial impedance configuration parameter to control the rehabilitation training institution to perform passive training on the target object's limbs along a preset desired movement trajectory.

[0032] Step S2: In the passive training phase, real-time state data and real-time torque data of the target object are collected, and active force characterization value is calculated based on the real-time torque data, and bilateral motion deviation value is calculated based on the real-time state data; Specifically, during the passive training phase, the joint angle values ​​output by the joint angle sensors of the target object's limbs, the real-time torque values ​​output by the drive motors of each joint, and the surface electromyography (EMG) signals output by the surface EMG sensors attached to the preset force-generating muscle groups of the limbs are synchronously acquired at a sampling frequency of 50Hz. After bandpass filtering of each EMG signal, its root mean square value in the time domain is calculated as the amplitude feature. The power spectrum is obtained by performing a Fourier transform on the filtered EMG signal. The frequency value corresponding to the cumulative energy reaching 50% in the power spectrum is calculated as the frequency feature. The amplitude feature and the frequency feature are used as EMG feature parameters.

[0033] The joint angle values, real-time torque values, and electromyographic (EMG) characteristic parameters at the same moment are combined by limb channel to construct a multimodal data frame sequence. The EMG characteristic parameters of each limb are extracted from the current data frame, input into a preset EMG-force mapping table, and the active force expression value of each limb at the current moment is obtained by looking up the table. This expression value is quantized into a value from 0 to 100, representing the patient's current active force expression level.

[0034] Extract the joint angle values ​​of the first and second limbs of each symmetrical limb pair from the current data frame. Calculate the absolute value of the difference between the joint angle values ​​of the first and second limbs at the same acquisition time as the bilateral motion deviation value of the symmetrical limb pair.

[0035] Step S3: Based on the active force characterization value and the bilateral motion deviation value, continuously adjust the resistance and trajectory of the rehabilitation training mechanism during the passive training phase; Specifically, during the continuous passive training phase, the current data frame is read sequentially from the constructed multimodal data frame sequence, the active force expression value of each limb in the data frame is extracted, and each active force expression value is compared with the preset expected force expression range, which is preset according to the current training phase.

[0036] When the active force exertion value of any limb falls below the lower limit of the desired force exertion range, a resistance reduction command is generated, and the resistance reduction step size is linearly calculated based on the difference between the active force exertion value and the lower limit. When the active force exertion value of any limb exceeds the upper limit of the desired force exertion range, a resistance increase command is generated, and the resistance increase step size is linearly calculated based on the difference between the active force exertion value and the upper limit. When the active force exertion values ​​of all limbs fall within the desired force exertion range, the current resistance value is maintained. The generated resistance adjustment command and corresponding step size are output to the resistance execution unit of the rehabilitation training institution in real time, and the resistance adjustment is completed in the next control cycle.

[0037] Extract the bilateral motion deviation values ​​of each symmetrical limb pair from the current data frame, and compare each bilateral motion deviation value with a preset deviation tolerance threshold. When the bilateral motion deviation value of any symmetrical limb pair exceeds the deviation tolerance threshold, extract the joint angle values ​​of the first limb and the second limb in the symmetrical limb pair, and calculate the absolute value of the first angle difference between the joint angle value of the first limb and the first expected angle value at the corresponding time in the preset expected trajectory, and the absolute value of the second angle difference between the joint angle value of the second limb and the second expected angle value at the corresponding time in the preset expected trajectory.

[0038] The absolute values ​​of the first and second angle differences are compared. The limb with the larger absolute value of the angle difference is identified as the deviation-dominant side, and the other limb is identified as the deviation-following side. Based on the difference between the joint angle value of the deviation-dominant side and the preset desired trajectory, a trajectory compensation amount is generated according to a preset proportional coefficient. This trajectory compensation amount is superimposed on the drive command of the deviation-following side to correct the motion trajectory of the deviation-following side.

[0039] Step S4: Evaluate the completion rate of the real-time status data collected during the passive training phase. When the evaluation result reaches the phase transition standard, switch the rehabilitation training institution to the active assisted training mode to continue training until the training end instruction is received.

[0040] Specifically, during the continuous passive training phase, the collected real-time status data is used to assess the degree of motor completion. The actual movement trajectories of each limb at multiple consecutive acquisition times are extracted from the real-time status data collected during the passive training phase. These trajectories are compared with the expected trajectories at corresponding times in the preset expected trajectory, and the movement conformity deviation at each acquisition time is calculated. This deviation is the absolute value of the difference between the actual joint angle value and the expected joint angle value. The duration for which the movement conformity deviation at multiple consecutive acquisition times is less than a preset deviation threshold is counted. When this duration reaches the preset stage transition duration, the motor completion assessment result is deemed to have met the stage transition standard. A stage switching command is generated based on the stage transition standard, and an impedance adjustment factor matching the current real-time status data is extracted from the rehabilitation model library. Based on this impedance adjustment factor, the impedance configuration parameters are adjusted to the initial values ​​suitable for active assisted training. In response to the stage switching command, the rehabilitation training institution switches to the active assisted training mode with the adjusted impedance configuration parameters. Training continues in the active assisted training mode, and real-time status data and real-time torque data are continuously collected. The dynamic adjustment process of resistance adjustment and trajectory calibration is repeatedly executed until a training end command is received.

[0041] This invention provides a rehabilitation training method that simultaneously collects real-time status data and real-time torque data of the target subject, and calculates the active force expression value based on the real-time torque data. This achieves dynamic matching between training resistance and the patient's real-time force expression ability, solving the problem of training intensity being out of sync with the patient's actual ability due to lag in resistance adjustment in existing technologies, and avoiding insufficient training or secondary injury. By calculating bilateral movement deviation values ​​based on real-time status data, and continuously adjusting the trajectory of the rehabilitation training mechanism based on these deviation values ​​during the passive training phase, adaptive calibration of the limb coordinated movement trajectory is achieved, effectively suppressing excessive compensatory behavior of the healthy limb and ensuring that the affected limb receives sufficient training load. By evaluating the movement completion rate of the real-time status data collected during the passive training phase, and switching to active assisted training mode when the evaluation result reaches the stage transition standard, continuous and fine-grained data collection and evaluation of the training process are achieved. This provides therapists with traceable and quantifiable evidence of rehabilitation progress, supporting the refined adjustment of training programs.

[0042] like Figure 3 As shown, in one embodiment, obtaining initial state data of the target object and obtaining impedance configuration parameters from the rehabilitation training institution based on the initial state data includes: The rehabilitation training institution collects the initial joint angle values ​​of the target object's limbs, as well as the initial surface electromyography (EMG) signals output by the surface EMG sensors attached to the preset force-generating muscle groups of the target object's limbs. Among them, the preset force-generating muscle groups refer to the specific muscle groups that are pre-set for collecting surface electromyography signals, including the main force-generating muscle positions of the limbs such as the bilateral biceps brachii and quadriceps femoris. The surface electromyography sensors are attached to the skin surface corresponding to these muscle groups.

[0043] Initial surface electromyography (EMG) signal: refers to the original EMG signal collected by surface EMG sensors attached to the preset muscle groups at the start of training. This signal reflects the electrophysiological activity of the muscles at rest or in the initial state.

[0044] Initial joint angle values: These refer to the initial position angle data of each joint of the target object's limbs collected by the joint angle sensors of the rehabilitation training institution when training begins, serving as the position reference before passive training begins.

[0045] Specifically, at the start of training, the rehabilitation training institution simultaneously collects the initial joint angle values ​​of the target subject's left upper limb, right upper limb, left lower limb, and right lower limb using angle sensors configured at each joint, recording the posture angle of each joint in the initial position. Simultaneously, surface electromyography (EMG) sensors, pre-attached to the skin surface of pre-defined muscle groups such as the bilateral biceps and quadriceps muscles of the target subject, collect the initial surface EMG signals of each muscle group. These signals are output in analog voltage form and converted from analog to digital to form a raw digital EMG data stream.

[0046] The initial surface electromyography (EMG) signals are filtered and feature extracted to obtain initial EMG feature parameters. Specifically, the initial surface electromyography (EMG) signals from each limb channel were bandpass filtered using a 20Hz to 450Hz bandpass filter to remove power frequency interference and low-frequency motion artifact noise. The root mean square (RMS) value in the time domain was calculated for the filtered signal. This RMS value was obtained by taking the square root of the mean of the squared signal amplitude within a time window, serving as the amplitude characteristic reflecting muscle exertion intensity. Simultaneously, a Fourier transform was performed on the filtered signal to convert the time-domain signal to the frequency domain, obtaining the power spectrum distribution. When the accumulated energy in the power spectrum reached 50% of the total energy, the frequency value corresponding to this accumulation point was extracted as the median frequency, which serves as the frequency characteristic reflecting muscle fatigue state. The amplitude and frequency characteristics of the same limb channel were combined to form the initial EMG characteristic parameters for that limb.

[0047] The initial joint angle values ​​of each limb are integrated with the initial electromyographic characteristic parameters to obtain the initial state data; Specifically, the initial joint angle values ​​and initial electromyographic (EMG) feature parameters of the same limb channel are concatenated according to a preset data structure to form the initial state sub-vector of that limb. Taking the left upper limb as an example, the initial joint angle values ​​collected for this limb are combined with the amplitude and frequency features in the EMG feature parameters of the left upper limb into a three-dimensional sub-vector. Traversing all limb channels, the initial state sub-vectors of the four limbs—left upper limb, right upper limb, left lower limb, and right lower limb—are arranged in a fixed order and merged to form initial state data containing information from all limb channels. This initial state data is stored in the form of a multi-dimensional vector and serves as the complete input for subsequent matching rehabilitation stages.

[0048] Based on the initial state data, the rehabilitation stage in the rehabilitation training institution is matched to obtain the rehabilitation stage to which the target object belongs, and the impedance configuration parameters of the rehabilitation stage are extracted.

[0049] The rehabilitation stage refers to the training stages pre-divided according to the patient's rehabilitation process, including the initial passive training stage and the intermediate active assisted training stage. Each stage corresponds to a preset expected exertion range and impedance configuration parameters.

[0050] Impedance configuration parameters: These refer to the resistance settings used by the rehabilitation training institution during training. They are used to control the torque output of the drive motor and determine the intensity of resistance that the patient must overcome during exercise.

[0051] The rehabilitation model library is divided into multiple discrete rehabilitation stages according to the patient's rehabilitation process. Each stage corresponds to a set of standard state vectors and impedance configuration parameters adapted to that stage. The standard state vector is composed of the expected joint angle values ​​and expected electromyographic characteristic parameters of each limb channel under the ideal state of that stage. The expected joint angle values ​​reflect the movement trajectory positions that the limbs should achieve in that stage, and the expected electromyographic characteristic parameters include the preset amplitude and frequency characteristics of that stage, used to characterize the muscle exertion ability and fatigue tolerance level that the patient should possess in that stage. The impedance configuration parameters are the resistance settings used by the rehabilitation training institution when performing training in that stage, determining the magnitude of the torque output by the drive motor. This parameter forms a correlation mapping relationship with the standard state vector.

[0052] Specifically, the initial state data is used as a query vector to traverse the rehabilitation model library pre-stored in the rehabilitation training institution's storage unit. This rehabilitation model library contains multiple rehabilitation stages, each associated with a standard state vector and corresponding impedance configuration parameters. The standard state vector consists of the joint angle values ​​and electromyographic characteristic parameters under the ideal state of that stage. The Euclidean distance between the initial state data and each standard state vector is calculated. This Euclidean distance is obtained by taking the square root of the sum of the squares of the differences between the corresponding dimensions of the two vectors, and is used to quantify the similarity between the initial state and the standard state of each rehabilitation stage. After traversing all rehabilitation stages, the rehabilitation stage corresponding to the standard state vector with the smallest Euclidean distance is selected as the rehabilitation stage to which the target object belongs. The corresponding impedance configuration parameters are extracted from the record of this stage and output as the initial resistance setting value for subsequent passive limb training to the resistance execution unit of the rehabilitation training institution.

[0053] The method provided in this embodiment synchronously acquires initial joint angle values ​​and initial surface electromyography (EMG) signals, and performs filtering and feature extraction on the EMG signals to obtain EMG feature parameters. Integrating these two data points forms initial state data, providing an objective basis for matching impedance configuration parameters by fusing multi-source information. This avoids relying on the therapist's subjective experience to set initial resistance, ensuring that the initial training intensity is accurately matched to the patient's actual ability. By extracting the root mean square value in the time domain as the amplitude feature and the median frequency in the frequency domain as the frequency feature from the surface EMG signals, the amplitude feature reflecting muscle exertion intensity and the frequency feature reflecting muscle fatigue state are combined as EMG feature parameters. This quantifies the patient's initial muscle function state from both the dimensions of exertion ability and fatigue tolerance, making the matching results more closely reflect the patient's actual physiological condition. By performing similarity matching between the initial state data and the standard state vectors of each rehabilitation stage in the rehabilitation model library, the impedance configuration parameter corresponding to the standard state vector with the highest matching degree is selected as the initial impedance configuration parameter. This achieves automatic association between the initial training resistance and the patient's rehabilitation stage, providing a precise starting point for subsequent dynamic closed-loop control.

[0054] In one embodiment, controlling the rehabilitation training institution to perform passive limb training on the target object based on the impedance configuration parameters includes: Read the matched impedance configuration parameters, and determine the expected movement trajectory of each limb based on the impedance configuration parameters. The expected movement trajectory includes the angle sequence of each joint over time. Specifically, the impedance configuration parameter is read. This parameter, stored in numerical form, is used to calibrate the base resistance level for the current training phase. Based on this impedance configuration parameter, the corresponding expected movement trajectory is selected from the rehabilitation model library. This expected movement trajectory records the sequence of joint angle values ​​over time in the form of discrete sampling points. Each sampling point includes a timestamp and the corresponding angle values ​​for the left upper limb shoulder joint, elbow joint, right upper limb shoulder joint, elbow joint, left lower limb hip joint, knee joint, and right lower limb hip joint. The rehabilitation model library pre-stores multiple expected movement trajectories corresponding to different resistance levels. The higher the resistance level, the more the trajectory amplitude and movement speed are adjusted to adapt to the needs of different training intensities.

[0055] The expected movement trajectory is converted into drive commands for each joint drive motor, and the drive commands include the target position and target velocity at each moment; Specifically, the expected movement trajectory is sampled according to the control cycle. At the beginning of each control cycle, the expected angle values ​​of each joint at the current moment are extracted from the expected movement trajectory as the target position. The target speed is calculated by the position difference between adjacent sampling points and the sampling time interval. The calculation formula is: target speed equals the expected angle difference between two adjacent sampling points divided by the sampling time interval. For each joint drive motor, the calculated target position and target speed are encapsulated into a drive command data packet. This data packet contains a joint identifier, target position value, target speed value, and timestamp information. This conversion process is repeated within each control cycle to ensure that the drive command can track changes in the expected movement trajectory in real time.

[0056] The drive command is sent to the joint drive motors of the rehabilitation training institution, and the drive motors drive the limbs of the target object to perform passive movement along the expected movement trajectory according to the drive command.

[0057] Specifically, the drive command data packets are sent to the corresponding joint drive motor controllers in the rehabilitation training institution via a real-time communication bus. The communication bus adopts a high-speed fieldbus protocol to ensure the real-time and synchronous transmission of commands, with drive commands for all joint motors being sent synchronously within the same control cycle. After receiving the commands, the drive motor controllers parse the target position and target speed, and adjust the motor output torque and speed through an internal closed-loop control algorithm, causing the motor shaft to drive the transmission mechanism, thereby driving the patient's corresponding limb joints to move according to the target position and target speed. The drive motors of each joint work together to drive the patient's limbs to perform passive movements along the expected movement trajectory. During the movement, the angle sensors of each joint provide real-time feedback on the actual position, forming a closed-loop verification with the target position in the drive command to ensure the accurate execution of the movement trajectory.

[0058] The method provided in this embodiment automatically associates impedance configuration parameters with movement trajectories by reading the matched impedance configuration parameters and determining the expected movement trajectory of each limb accordingly. This avoids the subjective bias of therapists manually setting movement trajectories and ensures that the movement trajectory during passive training is accurately matched with the patient's current training intensity. By converting the expected movement trajectory into drive commands for the joint drive motors and sending these commands to the respective joint drive motors, the synchronous generation and distribution of multi-joint movement commands are achieved. This ensures that the joints of the limbs move in coordination according to a unified time reference, overcoming the problem of poor limb coordination in existing technologies. The drive motors, following the drive commands, drive the patient's limbs to perform passive movements along the expected movement trajectory. This allows the affected limb to receive stable and repetitive movement stimulation under the guidance of a preset trajectory, establishing a standard movement pattern foundation for subsequent active assisted training and ensuring the standardization and safety of the rehabilitation training process.

[0059] like Figure 4 As shown, in one embodiment, during the passive training phase, real-time state data and real-time torque data of the target object are collected, and an active force characterization value is calculated based on the real-time torque data. The bilateral motion deviation value is also calculated based on the real-time state data, including: The real-time status data collected by the rehabilitation training institution during the passive training phase is obtained, and the motion trajectory parameters of the first limb and the second limb are extracted from the real-time status data. Among them, the first limb and the second limb refer to the left and right limbs that constitute a symmetrical limb pair. For example, the left upper limb and the right upper limb constitute a symmetrical upper limb pair, and the left lower limb and the right lower limb constitute a symmetrical lower limb pair. In each symmetrical limb pair, the first limb and the second limb can represent the left limb and the right limb, or they can be interchanged according to the actual configuration.

[0060] Specifically, during the continuous operation of the passive training phase, the rehabilitation training institution synchronously collects the joint angle values ​​output by the joint angle sensors at a sampling frequency of 50Hz. The collected data is then aligned with timestamps and stored in a data buffer, forming a real-time status data stream. For each sampling moment, a data frame is read from the buffer. This data frame contains the angle values ​​of the left upper limb shoulder joint, left upper limb elbow joint, right upper limb shoulder joint, right upper limb elbow joint, left lower limb hip joint, left lower limb knee joint, right lower limb hip joint, and right lower limb knee joint. The left and right upper limbs are defined as a symmetrical upper limb pair, and the left and right lower limbs are defined as a symmetrical lower limb pair.

[0061] For symmetrical upper limb pairs, the joint angle values ​​of the left upper limb are extracted from the current data frame as the motion trajectory parameters of the first limb, and the joint angle values ​​of the right upper limb are extracted as the motion trajectory parameters of the second limb. For symmetrical lower limb pairs, the joint angle values ​​of the left lower limb are extracted as the motion trajectory parameters of the first limb, and the joint angle values ​​of the right lower limb are extracted as the motion trajectory parameters of the second limb. The extracted motion trajectory parameters are stored in the form of four-dimensional vectors, corresponding to the four limb channel data of the two symmetrical limb pairs.

[0062] For each symmetrical limb pair, the trajectory deviation value between the motion trajectory parameters of the first limb and the second limb at the same acquisition time is calculated, and the trajectory deviation values ​​at multiple consecutive acquisition times are averaged to obtain the bilateral motion deviation value of each symmetrical limb pair. Among them, the bilateral movement deviation value refers to the quantitative index obtained after mean processing, which represents the consistency of bilateral movement of symmetrical limb pairs over a period of time. The smaller the value, the better the bilateral movement coordination.

[0063] Specifically, for each symmetrical limb pair, the joint angle values ​​of the first limb at the same acquisition time are read from the motion trajectory parameters extracted in step one. Joint angle value with the second limb Calculate the trajectory deviation of the symmetrical limb pair at that moment. The calculation formula is: The calculation is performed independently in the symmetrical upper limb pair and the symmetrical lower limb pair, and the trajectory deviation values ​​of the upper limb and the lower limb are obtained respectively.

[0064] The calculated trajectory deviation value for each symmetrical limb pair They are stored sequentially into the corresponding deviation buffer queues according to time order, and the window length of each buffer queue is preset to N sampling times (N=100).

[0065] Once the buffer queue is filled with N trajectory deviation values, the arithmetic mean of all deviation values ​​in the queue is calculated to obtain the bilateral motion deviation value. : , This mean is used as the bilateral motion deviation value corresponding to the current window.

[0066] Bilateral motor deviation values ​​are stored separately as upper limb bilateral motor deviation values ​​and lower limb bilateral motor deviation values, used to characterize the degree of coordination of bilateral limb movements over a recent period. After the mean is processed, the buffer queue slides forward, removing the oldest deviation value and adding the latest deviation value, and then proceeds to the next round of calculation.

[0067] The real-time torque data output by the drive motors of each joint in the rehabilitation training institution is obtained, and the difference between the real-time torque value at each acquisition time and the preset passive training reference torque value is calculated to obtain the torque deviation value at each acquisition time. The preset passive training baseline torque value refers to the baseline torque required for the drive motor to move the patient's limb along a preset trajectory when the patient is completely passive and does not actively exert force in passive training mode. This value is obtained through pre-calibration and serves as a reference benchmark for judging whether the patient is actively exerting force. For example, taking the right upper limb elbow joint as an example, baseline calibration is performed before passive training begins. The target subject's right upper limb is placed in a completely relaxed state without any active force applied. The rehabilitation training mechanism is controlled to move the right upper limb elbow joint from the extended position (0 degrees) to the 90-degree position at a low speed (e.g., 5 degrees / second) and then back to the starting position at a constant speed. During this process, the drive motor needs to output a certain torque value to overcome joint friction, soft tissue resistance, and the limb's own weight. The real-time torque values ​​corresponding to each angle position during the entire movement are recorded at a sampling frequency of 10Hz, forming an "angle-torque" mapping table. For example, when the elbow joint is flexed to 45 degrees, the drive motor outputs a torque of 2.5 N·m; when flexed to 90 degrees, the output torque is 3.2 N·m. The values ​​in this mapping table represent the passive training baseline torque values ​​for the joint at different angular positions. In actual training, if the real-time torque value measured at the same angular position (e.g., 45 degrees) is 1.8 N·m, lower than the baseline value of 2.5 N·m, it indicates that the patient actively exerted force to assist movement at that moment, resulting in a negative torque deviation; if the real-time torque value is 3.5 N·m, higher than the baseline value, it indicates that the patient is engaging in antagonistic movement. By calculating the difference between the real-time torque value and the baseline value, the degree of active force exertion by the patient can be quantified.

[0068] Specifically, during the passive training phase, the controllers of the drive motors for each joint in the rehabilitation training institution collect the output torque values ​​of the motors in real time at a sampling frequency of 50Hz, and upload the torque data to the control unit via the communication bus. After being aligned with the joint angle data at the same timestamp, the data is stored in the data cache area.

[0069] For each acquisition moment, the control unit reads the real-time torque value Ta of each joint from the buffer. Simultaneously, based on the current angular position of each joint, it looks up the corresponding passive training reference torque value Tb from a pre-stored "angle-reference torque" mapping table. The mapping table stores reference torque values ​​at discrete angle points (e.g., one sampling point every 5 degrees). For non-sampling point angles, a linear interpolation method is used to calculate the corresponding reference value. After reading the real-time torque value and the reference value, the torque deviation ΔT = Ta is calculated. Tb is measured in Newton-meters (N·m). This calculation is performed independently on each joint's drive motor channel to obtain the torque deviation value for each joint at the current moment. The torque deviation value can be positive or negative: when ΔT < 0, it indicates that the patient is actively exerting force to assist movement, and the drive motor's output torque is lower than the baseline; when ΔT > 0, it indicates that the patient is resisting movement or there is abnormal resistance; when ΔT = 0, it indicates that the patient is completely passive and has no active force exertion.

[0070] The torque deviation values ​​at multiple consecutive acquisition times are accumulated to obtain the active force characterization value.

[0071] Specifically, for each joint channel, a cumulative window length M is set (e.g., M=100, corresponding to a 2-second time window). The torque deviation values ​​ΔTi of the most recent M acquisition moments are retrieved from the torque deviation value cache queue. All deviation values ​​within this window are summed to obtain the active force characterization value A for that joint channel. The calculation formula is as follows: .

[0072] The accumulation operation is performed in the integer or floating-point field. The sign and magnitude of the accumulation result reflect the net contribution of the joint to active force exertion within the window: if the accumulated value is negative and the absolute value is large, it indicates that the patient has been continuously exerting active force during that time period; if the accumulated value is positive, it indicates that the patient is exerting counterforce or that there is abnormal resistance in the system. The active force characterization value A is expressed in Newton-meter-seconds (N·m·s) and serves as the basis for subsequent dynamic resistance adjustment.

[0073] If the value is lower than the preset lower limit of the expected force range, the system determines that the patient's force is insufficient and generates a resistance reduction command; if it is higher than the upper limit, it determines that the force is too strong and generates a resistance increase command. After the cumulative calculation is completed, the cache queue slides forward, removing the oldest deviation value and adding the latest deviation value, repeating the above process to achieve continuous updating of the active force representation value. This calculation is performed independently in each joint channel, but the resistance adjustment decision needs to take into account the active force representation values ​​of all limbs, based on the patient's overall force state.

[0074] The method provided in this embodiment extracts the motion trajectory parameters of symmetrical limb pairs from real-time status data, calculates the trajectory deviation value of both limbs at the same acquisition time, and averages the deviation values ​​at multiple consecutive acquisition times to obtain the bilateral motion deviation value, providing a stable and reliable deviation basis for subsequent dynamic trajectory calibration. By acquiring the real-time torque data output by the drive motors of each joint, the difference between the real-time torque value at each acquisition time and the preset passive training benchmark torque value is calculated to obtain the torque deviation value, which can quantify the instantaneous deviation between the patient's active force and the benchmark value. By accumulating the torque deviation values ​​at multiple consecutive acquisition times, an active force characterization value is obtained, providing a quantitative basis reflecting the patient's overall force contribution for dynamic resistance adjustment.

[0075] like Figure 5 As shown, in one embodiment, the step of continuously and dynamically adjusting the resistance and trajectory of the rehabilitation training mechanism during the passive training phase based on the active force characterization value and the bilateral motion deviation value includes: Each of the active force exertion characterization values ​​is compared with a preset desired force exertion range. When the active force exertion characterization value deviates from the desired force exertion range, the corresponding deviation state is obtained, and a resistance adjustment control signal is generated based on the deviation state. The preset expected exertion range refers to a reasonable range of active exertion performance values ​​pre-defined based on the patient's current rehabilitation stage, including an upper and lower limit. This range is pre-configured by the therapist based on the patient's rehabilitation progress and stored in the control unit. It is used to determine whether the patient's current active exertion level is in an ideal state. When the active exertion performance value is below the lower limit, it is considered insufficient exertion; when it is above the upper limit, it is considered excessive exertion; and when it falls within the range, it is considered moderate exertion.

[0076] Deviation state: refers to the degree of difference between the active force performance value and the expected force range, specifically including two situations: negative deviation below the lower limit and positive deviation above the upper limit. Deviation state is described by both the direction of deviation (negative or positive) and the magnitude of deviation (difference).

[0077] Specifically, during the continuous operation of the passive training phase, the control unit reads the active force representation values ​​calculated from each joint channel from the data buffer. For each joint channel, a preset expected force range is read, which includes a lower threshold and an upper threshold, pre-set and stored in the control unit according to the patient's current rehabilitation stage. The active force representation value is compared with the expected force range: if the active force representation value is lower than the lower threshold, it is determined to be a deviation state of insufficient force, and a resistance reduction command is generated; if the active force representation value is higher than the upper threshold, it is determined to be a deviation state of excessive force, and a resistance increase command is generated; if the active force representation value falls within the range, no resistance adjustment command is generated. For deviation states of insufficient or excessive force, the adjustment step size is further calculated based on the deviation of the active force representation value from the range boundary; the larger the deviation, the larger the adjustment step size. The adjustment direction and adjustment step size information are encapsulated into a resistance adjustment control signal, which contains a joint identifier, an adjustment direction identifier (upward or downward), and a step size value in the form of a data packet, ready to be sent to the resistance execution unit.

[0078] The resistance adjustment control signal is output to the resistance execution unit of the rehabilitation training institution to perform resistance adjustment; Among them, the resistance actuator refers to the hardware component in the rehabilitation training institution responsible for performing resistance adjustment. It usually includes a magnetic powder brake, a servo motor torque control module or an adjustable damping device. After receiving the resistance adjustment control signal, it completes the dynamic adjustment of the resistance value in the next control cycle.

[0079] Specifically, the resistance adjustment control signal line is sent to the corresponding resistance execution unit in the rehabilitation training institution. Upon receiving the signal, the resistance execution unit parses the joint indicator, adjustment direction indicator, and step length value. If the command is to increase resistance, the unit increases the current resistance value based on the step length value, increasing the output torque of the drive motor; if the command is to decrease resistance, the unit decreases the current resistance value based on the step length value, reducing the output torque of the drive motor. The resistance adjustment is completed within the next control cycle, and the adjusted resistance value takes effect immediately, influencing subsequent drive motor control. After completing the adjustment, the resistance execution unit returns an execution confirmation signal to the control unit, indicating that the resistance has been adjusted according to the command. This resistance adjustment process continues continuously within each control cycle based on the real-time changes in the active force characteristic value.

[0080] When any of the bilateral motion deviation values ​​is greater than a preset deviation tolerance threshold, the degree of deviation between the real-time status data corresponding to the bilateral motion deviation value and the preset joint trajectory is identified, and the deviation-dominant side and the deviation-following side are determined according to the degree of deviation. The preset deviation tolerance threshold refers to the upper limit of the allowable range of bilateral motion deviation values. When the bilateral motion deviation value exceeds this threshold, bilateral motion incoordination is determined, triggering a trajectory calibration operation. This threshold is preset based on clinical experience and individual patient conditions; for example, it is set to 5 degrees for the upper limbs and 8 degrees for the lower limbs.

[0081] The trajectory compensation amount of the deviation-following side is calculated based on the deviation-dominant side and the preset desired angle value, and the motion trajectory of the deviation-following side is corrected based on the trajectory compensation amount.

[0082] Specifically, after determining the deviation-dominant side and the deviation-following side, the actual joint angle value of the deviation-dominant side and the expected angle value at the corresponding moment in the preset joint trajectory are read, and the angle difference between the two is calculated as the deviation amount of the deviation-dominant side. This deviation amount reflects the degree to which the deviation-dominant side deviates from the expected trajectory.

[0083] The deviation is multiplied by a preset proportional coefficient to obtain the trajectory compensation amount. The proportional coefficient is set to 0.5 to ensure that the compensation amount is half of the deviation amount, avoiding overcorrection that could cause motion oscillations. The sign of the trajectory compensation amount is consistent with the deviation direction on the side dominated by the deviation, guiding the deviation to follow the actual motion trajectory of the side dominated by the lateral deviation.

[0084] The calculated trajectory compensation is added to the drive command of the deviation-following side. Specifically, the target position in the current drive command of the deviation-following side is read, and this target position is added to the trajectory compensation to generate a corrected target position, forming a new drive command. This corrected drive command is sent to the joint drive motor controller corresponding to the deviation-following side via a real-time communication bus. After receiving the corrected drive command, the drive motor controller moves the limb of the deviation-following side according to the new target position, making its movement trajectory approximate the actual trajectory of the deviation-dominant side, thereby achieving bilateral limb movement coordination. The calculation of trajectory compensation and command correction are repeated in each control cycle until the bilateral movement deviation value falls back to within the deviation tolerance threshold, forming a continuous closed-loop calibration mechanism.

[0085] The method provided in this embodiment adjusts resistance by comparing the active force expression value with the desired force range and generating a resistance adjustment signal. This dynamically matches the training resistance to the patient's real-time force state, avoiding the training intensity disconnect caused by fixed resistance and solving the problem of resistance adjustment lag. When bilateral movement deviation values ​​exceed a threshold, the method identifies the degree of deviation between real-time state data and preset joint trajectories, determining the dominant and follower sides of the deviation. This achieves precise localization of bilateral movement incoordination, providing a clear correction benchmark for trajectory calibration. By calculating the trajectory compensation amount of the follower side based on the dominant side and the preset desired angle value, and correcting its movement trajectory, the follower side moves closer to the actual trajectory of the dominant side, achieving adaptive and coordinated calibration of limb movement trajectories, suppressing compensatory behavior on the healthy side, and ensuring the affected side receives sufficient training load.

[0086] In one embodiment, when any of the bilateral motion deviation values ​​exceeds a preset deviation tolerance threshold, identifying the degree of deviation between the real-time state data corresponding to the bilateral motion deviation value and the preset joint trajectory, and determining the deviation-dominant side and the deviation-following side based on the degree of deviation, includes: When the bilateral motion deviation value is greater than the deviation tolerance threshold, locate the real-time status data corresponding to the bilateral motion deviation value, and extract the joint angle value of the first limb and the joint angle value of the second limb from the real-time status data. Specifically, during the continuous operation of the passive training phase, the bilateral motion deviation values ​​of each symmetrical limb pair are monitored in real time. When the bilateral motion deviation value of a certain symmetrical limb pair exceeds a preset deviation tolerance threshold, the trajectory calibration process is triggered. The control unit locates the acquisition time corresponding to the bilateral motion deviation value through the data index and extracts the data frame of that time from the real-time status data buffer. This data frame contains the joint angle values ​​of all limbs at that time and is stored separately by limb channel.

[0087] Based on the type of symmetrical limb pair being processed, the joint angle values ​​of the first limb and the second limb are extracted from the data frame. Taking an upper limb symmetrical limb pair as an example, the shoulder and elbow joint angle values ​​of the left upper limb are read from the data frame as the joint angle values ​​of the first limb, and the shoulder and elbow joint angle values ​​of the right upper limb are read as the joint angle values ​​of the second limb.

[0088] If the current lower limbs are symmetrical, the hip and knee angles of the left lower limb are read as the joint angles of the first limb, and the hip and knee angles of the right lower limb are read as the joint angles of the second limb.

[0089] Calculate the absolute value of the first angle difference between the joint angle value of the first limb and the first expected angle value at the corresponding moment in the preset joint trajectory; calculate the absolute value of the second angle difference between the joint angle value of the second limb and the second expected angle value at the corresponding moment in the preset joint trajectory. When the absolute value of the first angle difference is greater than the absolute value of the second angle difference, the first limb is determined to be the deviation-dominant side, and the second limb is determined to be the deviation-following side. Specifically, the absolute values ​​of the first and second angle differences are read and compared. If the absolute value of the first angle difference is greater than the absolute value of the second angle difference, it indicates that the actual movement of the first limb deviates from the preset expected trajectory more than that of the second limb. In this case, the first limb is identified as the deviation-dominant side, and the second limb is identified as the deviation-following side. This determination result is stored in the control unit as an identifier for subsequent trajectory compensation calculation. The deviation-dominant side indicates that the limb on that side is in a "lagging" or "overly" state in bilateral movement incoordination, and its actual trajectory serves as the calibration benchmark. The deviation-following side indicates that the actual trajectory of the limb on that side is relatively close to the expected trajectory and needs to converge towards the actual trajectory of the deviation-dominant side.

[0090] Conversely, the second limb is determined as the deviation-dominant side, and the first limb is determined as the deviation-following side.

[0091] Specifically, if the absolute value of the first angle difference is not greater than the absolute value of the second angle difference (i.e., the absolute value of the second angle difference is greater than the absolute value of the first angle difference), then the opposite determination is performed. The control unit determines the second limb as the deviation-dominant side and the first limb as the deviation-following side. If the absolute values ​​of the first and second angle differences are equal, either side is arbitrarily selected as the deviation-dominant side, or the determination result of the previous cycle is maintained to avoid trajectory oscillation caused by frequent switching. After the determination is completed, the identification information of the deviation-dominant side and the deviation-following side is transmitted to the trajectory compensation calculation module to generate the compensation amount to correct the motion trajectory of the deviation-following side. This determination process is repeated in each control cycle that triggers trajectory calibration to ensure that the identification of the deviation-dominant side and the deviation-following side is always based on the actual motion state at the current moment.

[0092] The method provided in this embodiment locates the real-time state data corresponding to bilateral motion deviation values ​​exceeding a threshold and extracts the joint angle values ​​of both limbs, accurately linking the deviation event with the specific time and the motion state of the limb, providing an accurate source of raw data for deviation analysis. By calculating the absolute value of the angle difference between the joint angle values ​​of both limbs and the corresponding expected angle values, the degree of deviation of both limbs from the preset trajectory is quantified into a comparable numerical index, achieving objective quantification of bilateral motion incoordination. By comparing the magnitude of the absolute value of the angle difference between the two sides, the side with the larger value is identified as the deviation-dominant side, and the side with the smaller value is identified as the deviation-following side, providing a clear calibration benchmark for the calculation of trajectory compensation, ensuring that the calibration action is accurately applied to the limb that needs correction.

[0093] In one embodiment, the assessment of the exercise completion rate of the real-time status data collected during the passive training phase, and the switching of the rehabilitation training institution to the active assisted training mode to continue training when the assessment result reaches the phase transition standard, until a training end instruction is received, includes: The motion trajectory data of multiple consecutive acquisition times are extracted from the real-time status data, and each of the motion trajectory data is compared with the preset expected motion trajectory to identify the motion matching deviation at each acquisition time. Specifically, during the continuous operation of the passive training phase, real-time status data frames are read sequentially from the data buffer at a sampling frequency of 50Hz. Each data frame contains the angle values ​​of each limb joint at the current acquisition time. For each data frame, the expected angle values ​​of each joint at the corresponding time are extracted from the preset expected motion trajectory. For each joint, the absolute value of the difference between the actual angle value and the expected angle value is calculated to obtain the motion conformity deviation of that joint at that acquisition time. For limbs containing multiple joints (such as the shoulder and elbow joints), the arithmetic mean of the deviations of each joint is calculated as the comprehensive motion conformity deviation of that limb. The upper limb deviation is taken as the average of the deviations of the shoulder and elbow joints, and the lower limb deviation is taken as the average of the deviations of the hip and knee joints. The calculated motion conformity deviations of each limb are stored in the evaluation buffer queue in the order of acquisition time, and the window length of each buffer queue is preset to 100 sampling times (corresponding to a duration of 2 seconds).

[0094] When the motion matching deviation is less than the preset deviation threshold at multiple consecutive acquisition times, the evaluation result is determined to have reached the stage transition standard. Specifically, the sequence of motion conformity deviations in the evaluation cache queue is monitored and evaluated in real time. Starting from the head of the queue and proceeding backwards, the number of sampling moments that consecutively meet the condition of "motion conformity deviation less than a preset deviation threshold" is counted. The preset deviation threshold is set to 5 degrees for the upper limbs and 8 degrees for the lower limbs. When the number of sampling moments that consecutively meet the condition reaches the preset window length (100 sampling moments), i.e., the motion conformity deviation of all sampling moments within 2 consecutive seconds is lower than the corresponding threshold, the motion completion assessment result is determined to have reached the stage transition standard. This determination process is performed independently for each limb; stage switching is only triggered when the assessment results of all limbs have reached the stage transition standard. This determination mechanism ensures that the patient can stably and continuously complete the movement along the preset expected motion trajectory during the passive training phase and is capable of entering the active assisted training phase. After the determination is completed, the moment when the stage transition standard is reached is recorded, and a stage transition marker is generated.

[0095] A phase switching instruction is generated based on the phase transition standard, and an impedance adjustment factor matching the current real-time status data is extracted from the rehabilitation training institution. Based on the impedance adjustment factor, the impedance configuration parameter is adjusted to an initial value that adapts to the active assisted training mode. Specifically, upon determining that the stage transition criteria have been met, a stage switching instruction is generated. This instruction includes the switching trigger time and the target training mode identifier (active assisted training mode). Simultaneously, real-time state data is read, including joint angle values ​​and electromyographic characteristic parameters for each limb. Using the current real-time state data as a query vector, the standard state vectors corresponding to each active assisted training stage in the rehabilitation model library are traversed. The Euclidean distance between the current state vector and each standard state vector is calculated, and the impedance adjustment factor corresponding to the standard state vector with the smallest Euclidean distance is selected. This impedance adjustment factor is a numerical coefficient used to adjust the impedance configuration parameters. The current impedance configuration parameters are read and multiplied by the impedance adjustment factor to obtain the initial impedance value adapted to the active assisted training mode. If the current impedance configuration parameter is R and the impedance adjustment factor is k, then the adjusted impedance configuration parameter is R×k.

[0096] In response to the stage switching command, the rehabilitation training institution is controlled to switch to the active assisted training mode with the adjusted impedance configuration parameters, and the training continues until a training end command is received.

[0097] Among them, the active assisted training mode refers to a training mode in which the patient actively exerts force to participate in the movement and the rehabilitation training institution provides auxiliary torque. Compared with the passive training mode, in this mode, the patient needs to actively complete most of the movement movements, and the rehabilitation training institution only provides necessary auxiliary support.

[0098] Specifically, the stage switching command is sent to the main controller of the rehabilitation training institution via a real-time communication bus. Upon receiving the stage switching command, the main controller parses the target training mode and the adjusted impedance configuration parameters. At the start of the next control cycle, the main controller switches the operating mode of the rehabilitation training institution from passive training mode to active assisted training mode, and simultaneously updates the impedance configuration parameters of the resistance actuator to the adjusted initial values. In active assisted training mode, the rehabilitation training institution provides corresponding auxiliary torque based on the real-time detection results of the patient's active exertion, allowing the patient to actively participate in the movement and lead the completion of the action. During training, real-time status data and real-time torque data are continuously collected, and the resistance and trajectory are continuously dynamically adjusted based on the active exertion characterization value and bilateral movement deviation value, forming a closed-loop control. This training process continues until an externally input training end command is received (e.g., a stop command sent by the therapist through the operating interface or the preset training time is reached), at which point the rehabilitation training institution stops the exercise and the training session ends.

[0099] The method provided in this embodiment extracts motion trajectory data from multiple consecutive acquisition times from real-time status data and compares it with a preset expected motion trajectory. It identifies the motion consistency deviation at each acquisition time, refining discrete statistical indicators into quantifiable data of trajectory consistency over a continuous time series, providing fine-grained quantifiable basis for rehabilitation progress assessment. When the motion consistency deviation at multiple consecutive acquisition times is less than a preset deviation threshold, the assessment result is determined to have reached the stage transition standard, establishing objective stage switching conditions based on continuous and stable motion performance, avoiding reliance on subjective experience to determine the switching timing. By generating a stage switching command and extracting an impedance adjustment factor matching the current real-time status data, the impedance configuration parameters are adjusted to an initial value suitable for the active assisted training mode before switching the training mode, achieving a smooth transition between passive and active assisted training, ensuring precise matching of training intensity and the patient's real-time motion ability.

[0100] like Figure 6 As shown, the present invention also provides a rehabilitation training device, applied to any of the rehabilitation training methods described above, comprising: The acquisition module is used to acquire initial state data of the target object, and obtain impedance configuration parameters from the rehabilitation training institution based on the initial state data, and control the rehabilitation training institution to perform passive limb training on the target object based on the impedance configuration parameters. The analysis module is used to collect real-time state data and real-time torque data of the target object during the passive training phase, calculate the active force characterization value based on the real-time torque data, and calculate the bilateral motion deviation value based on the real-time state data. The association module is used to continuously and dynamically adjust the resistance and trajectory of the rehabilitation training mechanism during the passive training phase based on the active force characterization value and the bilateral motion deviation value. The processing module is used to evaluate the exercise completion rate of the real-time status data collected during the passive training phase. When the evaluation result reaches the phase transition standard, the rehabilitation training institution is switched to the active assisted training mode to continue training until the training end instruction is received.

[0101] This invention provides a rehabilitation training device that simultaneously collects real-time status data and real-time torque data of the target object, and calculates the active force expression value based on the real-time torque data. This achieves dynamic matching between training resistance and the patient's real-time force expression ability, solving the problem of training intensity being out of sync with the patient's actual ability due to lag in resistance adjustment in existing technologies, and avoiding insufficient training or secondary injury. By calculating bilateral movement deviation values ​​based on real-time status data, and continuously adjusting the trajectory of the rehabilitation training device during the passive training phase based on these deviation values, adaptive calibration of the limb coordinated movement trajectory is achieved, effectively suppressing excessive compensation behavior of the healthy limb and ensuring that the affected limb receives sufficient training load. By evaluating the movement completion rate of the real-time status data collected during the passive training phase, and switching to active assisted training mode when the evaluation result reaches the phase transition standard, continuous and fine-grained data collection and evaluation of the training process are achieved. This provides therapists with traceable and quantifiable evidence of rehabilitation progress, supporting the refined adjustment of training programs.

[0102] This application also provides a computer device, such as... Figure 7 As shown, the computer device includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor. When the processor executes the computer program, it implements the steps in any of the above method embodiments, or when the processor executes the computer program, it implements the functions of each module / unit in the above device embodiments.

[0103] For example, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete this application. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the computer device.

[0104] Those skilled in the art will understand that Figure 7The computer device described is merely an example and does not constitute a limitation on the computer device. It may include more or fewer components than shown, or combine certain components, or different components. For example, the computer device may also include input / output devices, network access devices, buses, etc.

[0105] The aforementioned processor can be a Central Processing Unit (CPU), or other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), or Field Programmable Gate Arrays (FPGAs). Programmable Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0106] The memory can be an internal storage unit of the computer device, such as a hard drive or RAM. The memory can also be an external storage device of the computer device, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory can include both internal and external storage units of the computer device.

[0107] This application also provides a readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the various method embodiments above.

[0108] This application provides a computer program product that, when run on an electronic device, enables the electronic device to perform the steps described in the various method embodiments above.

[0109] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a photographing device / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0110] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0111] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0112] In the embodiments provided in this application, it should be understood that the disclosed apparatus / devices and methods can be implemented in other ways. For example, the apparatus / device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0113] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0114] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A rehabilitation training method, characterized in that, include: The initial state data of the target object is obtained, and impedance configuration parameters are obtained from the rehabilitation training institution based on the initial state data. Based on the impedance configuration parameters, the rehabilitation training institution is controlled to perform passive limb training on the target object. During the passive training phase, real-time state data and real-time torque data of the target object are collected, and active force characterization value is calculated based on the real-time torque data, and bilateral motion deviation value is calculated based on the real-time state data. Based on the active force characterization value and the bilateral motion deviation value, the resistance and trajectory of the rehabilitation training mechanism are continuously dynamically adjusted during the passive training phase; The real-time status data collected during the passive training phase is used to assess the degree of exercise completion. When the assessment result reaches the phase transition standard, the rehabilitation training institution is switched to active assisted training mode to continue training until a training end instruction is received.

2. The rehabilitation training method according to claim 1, characterized in that, The step of obtaining initial state data of the target object and obtaining impedance configuration parameters from the rehabilitation training institution based on the initial state data includes: The rehabilitation training institution collects the initial joint angle values ​​of the target object's limbs, as well as the initial surface electromyography (EMG) signals output by the surface EMG sensors attached to the preset force-generating muscle groups of the target object's limbs. The initial surface electromyography (EMG) signals are filtered and feature extracted to obtain initial EMG feature parameters. The initial joint angle values ​​of each limb are integrated with the initial electromyographic characteristic parameters to obtain the initial state data; Based on the initial state data, the rehabilitation stage in the rehabilitation training institution is matched to obtain the rehabilitation stage to which the target object belongs, and the impedance configuration parameters of the rehabilitation stage are extracted.

3. The rehabilitation training method according to claim 1, characterized in that, In the passive training phase, real-time state data and real-time torque data of the target object are collected, and active force characterization values ​​are calculated based on the real-time torque data. Bilateral motion deviation values ​​are also calculated based on the real-time state data, including: The real-time status data collected by the rehabilitation training institution during the passive training phase is obtained, and the motion trajectory parameters of the first limb and the second limb are extracted from the real-time status data. For each symmetrical limb pair, the trajectory deviation value between the motion trajectory parameters of the first limb and the second limb at the same acquisition time is calculated, and the trajectory deviation values ​​at multiple consecutive acquisition times are averaged to obtain the bilateral motion deviation value of each symmetrical limb pair. The real-time torque data output by the drive motors of each joint in the rehabilitation training institution is obtained, and the difference between the real-time torque value at each acquisition time and the preset passive training reference torque value is calculated to obtain the torque deviation value at each acquisition time. The torque deviation values ​​at multiple consecutive acquisition times are accumulated to obtain the active force characterization value.

4. The rehabilitation training method according to claim 1, characterized in that, The dynamic adjustment of resistance and trajectory of the rehabilitation training mechanism during the passive training phase based on the active force characterization value and the bilateral motion deviation value includes: Each of the active force exertion characterization values ​​is compared with a preset desired force exertion range. When the active force exertion characterization value deviates from the desired force exertion range, the corresponding deviation state is obtained, and a resistance adjustment control signal is generated based on the deviation state. The resistance adjustment control signal is output to the resistance execution unit of the rehabilitation training institution to perform resistance adjustment; When any of the bilateral motion deviation values ​​is greater than a preset deviation tolerance threshold, the degree of deviation between the real-time status data corresponding to the bilateral motion deviation value and the preset joint trajectory is identified, and the deviation-dominant side and the deviation-following side are determined according to the degree of deviation. The trajectory compensation amount of the deviation-following side is calculated based on the deviation-dominant side and the preset desired angle value, and the motion trajectory of the deviation-following side is corrected based on the trajectory compensation amount.

5. The rehabilitation training method according to claim 4, characterized in that, When any of the bilateral motion deviation values ​​exceeds a preset deviation tolerance threshold, the degree of deviation between the real-time state data corresponding to the bilateral motion deviation value and the preset joint trajectory is identified, and the deviation-dominant side and the deviation-following side are determined based on the degree of deviation, including: When the bilateral motion deviation value is greater than the deviation tolerance threshold, locate the real-time status data corresponding to the bilateral motion deviation value, and extract the joint angle value of the first limb and the joint angle value of the second limb from the real-time status data. Calculate the absolute value of the first angle difference between the joint angle value of the first limb and the first expected angle value at the corresponding moment in the preset joint trajectory; calculate the absolute value of the second angle difference between the joint angle value of the second limb and the second expected angle value at the corresponding moment in the preset joint trajectory. When the absolute value of the first angle difference is greater than the absolute value of the second angle difference, the first limb is determined to be the deviation-dominant side, and the second limb is determined to be the deviation-following side. Conversely, the second limb is determined as the deviation-dominant side, and the first limb is determined as the deviation-following side.

6. The rehabilitation training method according to claim 1, characterized in that, The process involves evaluating the completion rate of the real-time status data collected during the passive training phase. When the evaluation result reaches the phase transition standard, the rehabilitation training institution switches to active assisted training mode to continue training until a training end instruction is received. This includes: The motion trajectory data of multiple consecutive acquisition times are extracted from the real-time status data, and each of the motion trajectory data is compared with the preset expected motion trajectory to identify the motion matching deviation at each acquisition time. When the motion matching deviation is less than the preset deviation threshold at multiple consecutive acquisition times, the evaluation result is determined to have reached the stage transition standard. A phase switching instruction is generated based on the phase transition standard, and an impedance adjustment factor matching the current real-time status data is extracted from the rehabilitation training institution. Based on the impedance adjustment factor, the impedance configuration parameter is adjusted to an initial value that adapts to the active assisted training mode. In response to the stage switching command, the rehabilitation training institution is controlled to switch to the active assisted training mode with the adjusted impedance configuration parameters, and the training continues until a training end command is received.

7. A rehabilitation training device, characterized in that, The rehabilitation training method applied to any one of claims 1-6 includes: The acquisition module is used to acquire initial state data of the target object, and obtain impedance configuration parameters from the rehabilitation training institution based on the initial state data, and control the rehabilitation training institution to perform passive limb training on the target object based on the impedance configuration parameters. The analysis module is used to collect real-time state data and real-time torque data of the target object during the passive training phase, calculate the active force characterization value based on the real-time torque data, and calculate the bilateral motion deviation value based on the real-time state data. The association module is used to continuously and dynamically adjust the resistance and trajectory of the rehabilitation training mechanism during the passive training phase based on the active force characterization value and the bilateral motion deviation value. The processing module is used to evaluate the exercise completion rate of the real-time status data collected during the passive training phase. When the evaluation result reaches the phase transition standard, the rehabilitation training institution is switched to the active assisted training mode to continue training until the training end instruction is received.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the rehabilitation training method as described in any one of claims 1 to 6.

9. A readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the rehabilitation training method as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, enables the implementation of the steps of the rehabilitation training method as described in any one of claims 1 to 6.

Images