Testing Methods and Systems for Lower Limb Hip and Knee Training Equipment Based on Joint Movement Coordination

Abstract

This invention relates to the field of medical device technology, specifically a testing method and system for lower limb hip and knee training devices based on joint motion coordination. The method involves constructing an activation state recognition and transition map, indexing and identifying the actual muscle group activation state through this map to obtain the current motion mode during device training. Based on this current motion mode, the method transfers and adjusts the motion posture along the standard joint execution chain of the hip and knee joints in the lower limb, generating a standard motion coordination trajectory. Based on this standard motion coordination trajectory, the method determines whether the motion coordination of the device's simulated motion coordination trajectory is satisfactory, yielding a first test result. Based on the first test result, a specified motion scenario is arranged to perform a second zero-torque point test on the lower limb hip and knee training device, and the instantaneous control of gait torque is determined, yielding a second test result. This invention can accurately test and determine the motion coordination performance of lower limb hip and knee training devices, effectively optimizing device adaptation and assisting in the recovery of lower limb hip and knee joints in different patients.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, and in particular to a testing method and system for lower limb hip and knee training equipment based on joint movement coordination. Background Technology

[0002] With the increasing aging population and the growing number of patients with sports injuries and neurological injuries (such as stroke and spinal cord injury), rehabilitation training equipment is being used more and more widely in the field of medical rehabilitation. Lower limb hip and knee training equipment is mainly used to improve the motor function of the hip and knee joints, increase joint range of motion, strengthen muscles, and promote neuromuscular coordination. However, most existing lower limb hip and knee training devices only target the hip or knee joint individually, and cannot be tested by simulating the biomechanical characteristics of the coordinated movement of the hip and knee joints during actual gait. Furthermore, it is difficult to accurately test and analyze the instantaneous control of the coordination, rhythm, and gait torque between the hip and knee joints in patients with different injury degrees, making it difficult to accurately determine whether the equipment is qualified and affecting the actual training effect. Some lower limb hip and knee training devices only support single-function injury analysis, making it difficult to formulate corresponding training and testing strategies based on the specific injury status of the patient's lower limb hip and knee joints. This results in significant errors in the test results for different patients, hindering subsequent precise control optimization and improvement of the equipment. Summary of the Invention

[0003] This invention overcomes the shortcomings of the prior art and provides a testing method and system for lower limb hip and knee training equipment based on joint movement coordination.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0005] The first aspect of this invention provides a testing method for a lower limb hip and knee training device based on joint movement coordination, comprising the following steps:

[0006] S102: Based on the probability calculation of the correct movement mechanism in human anatomy, multidimensional injury data of the patient subject is used to constrain the tensor distribution of normal movement of the hip and knee joints of the lower limbs, so as to configure the device for a dedicated training strategy for the patient subject.

[0007] S104: Using a three-dimensional simulation motion model of the lower limb hip and knee device, perform coordinated injury training simulation on the patient, obtain muscle group activation simulation signals, analyze the trend evolution of the muscle group activation simulation signals, and obtain the actual muscle group activation status of the patient.

[0008] S106: Construct an activation state recognition and transfer map by combining different preset muscle group activation states with corresponding activation transfer indices. Index the actual muscle group activation status through the activation state recognition and transfer map to obtain the current movement mode during equipment training.

[0009] S108: Based on the current motion mode, the motion posture is transmitted and adjusted on the standard joint execution chain of the hip and knee of the human lower limb to generate a standard motion coordination trajectory. Based on the standard motion coordination trajectory, the motion coordination of the simulated motion coordination trajectory of the device is judged to determine whether the motion coordination is qualified, and the first test result is obtained.

[0010] S110: Based on the first test result, a specified motion scenario is arranged to perform a second test on the lower limb hip and knee training equipment at the zero torque point and to determine the instantaneous control of gait torque, thereby obtaining the second test result.

[0011] More specifically, step S102 includes the following steps:

[0012] Obtain patient case report information, extract multidimensional injury data of the patient's lower limb hip and knee joints from the case report information, and obtain the correct movement mechanism of the lower limb hip and knee joints based on big data network;

[0013] Based on multidimensional injury data, construct injury tensor matrices with different motion dimensions, calculate the covariance matrix of the injury tensor matrix, and obtain the constraint eigenvalues ​​and constraint eigenvectors of multidimensional injury data along different motion dimensions through the covariance matrix.

[0014] Based on the constraint feature value, the kernel function and bandwidth of the constraint feature vector are preset. The probability density estimate of the injury constraint that causes normal movement of the hip and knee joints of the lower limbs in different motion dimensions is calculated by using the kernel function and bandwidth.

[0015] A singular value decomposition algorithm is introduced. Based on the correct motion mechanism, the constraint feature vector of the injury tensor matrix is ​​decomposed in the singular value decomposition algorithm to obtain the core tensor of the injury degree in different motion dimensions and the factor matrix that the core tensor in different motion dimensions restricts the normal motion of the patient's lower limb hip and knee joints.

[0016] Obtain the patient's exclusive training task, preset the minimum task loss function according to the exclusive training task, update the core tensor and factor matrix based on the decomposition rank alternation constraint of the injury constraint probability density estimate until the minimum task loss function is reached, and generate the tensor distribution of the injury impact on the normal movement of the patient's lower limb hip and knee joints in different movement dimensions when the injury severity of the injury data generates multidimensional injury data.

[0017] The motion structure of the lower limb hip and knee training device is obtained. Based on the motion structure, the lower limb hip and knee joints of the patient are divided into N sub-joint process blocks. Training decision allocation is performed on each sub-joint process block based on the injury influence tensor distribution, and a special training strategy for the lower limb hip and knee training device for the patient is generated.

[0018] More specifically, step S104 includes the following steps:

[0019] The assembly design drawings of the lower limb hip and knee training equipment and the joint injury movement process after the patient's multidimensional injury data were obtained. Based on the assembly design drawings, a three-dimensional simulation motion model of the lower limb hip and knee equipment was constructed in SolidWorks modeling software. Based on the joint injury movement process, a joint coordination motion model of the lower limb hip and knee joint injury was constructed.

[0020] The joint coordination motion model of the patient's lower limb hip and knee joint after injury is simulated and tested by controlling the three-dimensional simulation running model to execute the exclusive training strategy. During the simulation test, electromyography technology is used to record the muscle activation status in sequence to obtain the simulated activation signals of multiple muscle groups of the patient.

[0021] Sine waveform spectra of each group of muscle group activation simulation signals are extracted. Based on the sinusoidal waveform spectra, a Hanning window function that balances the time-frequency resolution of muscle group activation is preset. The instantaneous autocorrelation function of the signal transition is calculated through the complex conjugate of the muscle group activation simulation signal. Based on the Hanning window function, the time delay parameter of the instantaneous autocorrelation function is transformed to obtain the time-frequency simulation distribution characterization of multiple groups of muscle group activation simulation signals output by the hip and knee joints of the patient's lower limbs.

[0022] The local maximum peak point and local minimum peak point of each muscle group activation simulation signal are detected and obtained. The upper and lower envelopes of the muscle group activation simulation signal are constructed by spline interpolation of the local maximum peak point and local minimum peak point. At this time, the balance value between the upper and lower envelopes is calculated.

[0023] A regeneration signal is output by removing the equalization value from multiple muscle group activation simulation signals. If the upper envelope of the regeneration signal is completely symmetrical with respect to the lower envelope, the intrinsic modes are continuously extracted from the signal according to the time-frequency simulation distribution characterization to form a time-frequency distribution intrinsic mode chain. The actual muscle group activation status of the patient during the lower limb hip and knee training device training is determined according to the time-frequency distribution intrinsic mode chain.

[0024] More specifically, step S106 includes the following steps:

[0025] Based on big data networks, we obtain ergonomic knowledge graphs and muscle group movement cases of the hip and knee joints of the lower limbs. We then extract several movement patterns when different preset muscle group activation states occur in the hip and knee joints of the lower limbs through the ergonomic knowledge graphs.

[0026] By extracting the historical movement trend span of each preset muscle group activation state that causes the lower limb hip and knee joint movements to produce a coordinated pattern through muscle group movement cases, activation transfer index of different preset muscle group activation states is preset based on the historical movement trend span.

[0027] Each preset muscle group activation state is defined as a state node, and the activation transition index is used as the index boundary between each state node and the corresponding movement pattern. An activation state recognition transition map is constructed by fitting the state nodes and the index boundary.

[0028] A trend change detection model is constructed, and the actual muscle group activation status of the patient during the training of the lower limb hip and knee training device is accumulated and calculated using the trend change detection model to obtain the current activation status accumulation sum.

[0029] If the current active state cumulative sum is detected to be greater than the preset cumulative threshold, then the trend change length of the current active state cumulative sum is obtained. Based on the current active state cumulative sum, the trend change length is weighted and assigned a smooth character to obtain the active state string about the state transition at different trend change nodes.

[0030] Based on the activation transition index, the next state of each character in the activation state string is identified by traversing the activation trend recognition transition map. The acceptance coefficient of the activation trend recognition transition map for each character is output. If the acceptance coefficient is greater than the preset acceptance coefficient, the current movement mode of the patient when using the lower limb hip and knee training device is determined according to the index boundary of the current character state.

[0031] More specifically, step S108 includes the following steps:

[0032] Based on big data networks, standard joint structure diagrams and joint activity mechanisms of the human lower limb hip and knee are obtained. Using the joint activity mechanisms as execution constraints, standard joint execution chains of the human lower limb hip and knee are constructed based on the activities of the standard joint structure diagrams on the execution constraints.

[0033] Based on multidimensional injury data, the injury characteristics of the patient's lower limb hip and knee joints are obtained. The injury characteristics and the current movement pattern are imported into the ergonomic knowledge graph for identification. The maximum extension landing point of the patient's maximum movement under the current movement pattern and the free movement width between adjacent joint connection structures in the standard joint structure diagram are output.

[0034] The first and last joints closest to the limit of extension on the standard joint execution chain are separated, and the root node of the first joint and the end effector of the end joint are constructed. Based on the preset limit of the transmission length threshold of the free range of motion, the end effector is moved to the limit of extension starting from the root node.

[0035] If the current transmission length between adjacent joint connection structures exceeds the limit transmission length threshold, the motion pose between adjacent joint connection structures is adjusted until the current transmission length between adjacent joint connection structures does not exceed the limit transmission length threshold.

[0036] Repeat the above steps to generate a standard motion coordination trajectory of the patient in the current motion mode under the condition of lower limb hip and knee joint injury. Based on the standard motion coordination trajectory, determine the motion rhythm and coordination of the lower limb hip and knee training device on the simulated motion coordination trajectory to obtain the first test result.

[0037] More specifically, the process of repeating the above steps to generate a standard motion coordination trajectory of the patient's lower limb hip and knee joint under the current motion mode, and then judging the motion rhythm and coordination of the lower limb hip and knee training device based on the standard motion coordination trajectory to obtain the first test result, specifically includes the following steps:

[0038] The arrival error tolerance of the preset limit extension landing point is repeated, and the above steps of moving the end effector to the limit extension landing point and adjusting the motion posture between adjacent joint connection structures are repeated to obtain the current arrival distance of the end effector relative to the limit extension landing point.

[0039] If the current arrival distance is less than the arrival error tolerance, stop the motion posture adjustment operation, generate the standard motion coordination trajectory of the patient in the current motion mode under the condition of lower limb hip and knee joint injury, and obtain the simulated motion coordination trajectory of the patient during the training process of the lower limb hip and knee training device.

[0040] By extracting key collaborative nodes of the lower limb knee joint following the joint activity mechanism from the ergonomic knowledge graph, the entropy weight algorithm is introduced to calculate the approximate entropy of each key collaborative node on the simulated motion collaborative trajectory compared to the standard motion collaborative trajectory, resulting in multiple sparse approximate entropies and multiple dense approximate entropies.

[0041] Calculate the ratio of the number of key collaborative nodes between sparse approximation entropy and dense approximation entropy, and determine whether the ratio is greater than a preset ratio.

[0042] If the quantity ratio is less than the preset quantity ratio, the lower limb hip and knee training equipment is calibrated as qualified equipment; if the quantity ratio is greater than the preset quantity ratio, the lower limb hip and knee training equipment is calibrated as unqualified equipment, and the first test result is obtained.

[0043] More specifically, step S110 includes the following steps:

[0044] If the first test result shows that the lower limb hip and knee training equipment is qualified, then a specified motion scenario is set up in the simulation process to test the lower limb hip and knee training equipment and obtain multiple sets of joint force sensing test parameters of the lower limb hip and knee training equipment.

[0045] By introducing Newton's laws of dynamics to construct a dynamic topology model, and using multiple sets of joint force sensing test parameters, the reaction force between the lower limb hip-knee movement gait and the specified movement scenario is calculated within the dynamic topology model. This yields the simulated zero-moment point trajectory of the patient's movement gait under the action of the lower limb hip and knee training device during the simulation process.

[0046] Obtain the expected test requirements of the lower limb hip and knee training device and the drive control base station, and plan the expected zero-moment point trajectory of the lower limb hip and knee training device to apply to the patient under the premise of a specified motion scenario based on the expected test requirements;

[0047] Based on the preset gait torque stations of the drive control base station, the deviation between the simulated zero torque point trajectory and the expected zero torque point trajectory on each gait torque station is calculated, and multiple misalignment deviation values ​​are obtained. Only the gait torque stations with misalignment deviation values ​​greater than the preset misalignment deviation threshold are extracted and defined as suspicious gait torque stations.

[0048] Construct a transient peak fluctuation domain, fit suspicious gait torque sites within the transient peak fluctuation domain, and output the actual transient peak fluctuation tomography triggered by suspicious gait torque during the simulation of the lower limb hip and knee training device;

[0049] If the actual instantaneous peak fluctuation fault is at least one level higher than the preset instantaneous peak fluctuation fault, it indicates that the instantaneous torque control performance of the lower limb hip and knee training device is poor, resulting in gait coordination error, and the second test result is obtained.

[0050] The second aspect of the present invention provides a testing system for a lower limb hip and knee training device based on joint motion coordination. The testing system includes a memory and a processor. The memory stores a testing method program for a lower limb hip and knee training device based on joint motion coordination. When the testing method program for a lower limb hip and knee training device is executed by the processor, the steps of the testing method for a lower limb hip and knee training device as described in any one of the present invention are implemented.

[0051] This invention addresses the technical deficiencies in the prior art, and its beneficial technical effects are as follows:

[0052] Based on the probability calculation of the correct movement mechanism in human anatomy, multidimensional injury data of the patient is used to constrain the tensor distribution of normal movement of the hip and knee joints of the lower limbs, so as to configure a dedicated training strategy for the patient. A three-dimensional simulation motion model of the lower limb hip and knee equipment is used to simulate coordinated injury training for the patient, acquiring simulated muscle activation signals. The trend evolution of these simulated muscle activation signals is analyzed to obtain the actual muscle activation state of the patient. Different preset muscle activation states are combined with corresponding activation transfer indices to construct an activation state recognition transfer map. The actual muscle activation state is indexed and identified using this map to obtain the current movement mode during equipment training. Based on the current movement mode, the movement posture is adjusted along the standard joint execution chain of the hip and knee of the lower limbs, generating a standard motion coordination trajectory. The motion coordination of the simulated motion coordination trajectory of the equipment is judged to determine its suitability, resulting in a first test result. Based on the first test result, a specified movement scenario is arranged to conduct a second zero-torque point test on the lower limb hip and knee training equipment, and the instantaneous control of gait torque is judged, resulting in a second test result. This invention can accurately test and determine whether lower limb hip and knee training equipment is qualified for the instantaneous control of movement coordination, rhythm and gait of injured patients, thereby providing a reliable basis for the subsequent optimization of lower limb hip and knee training equipment and effectively improving the equipment's adaptability and assisting the recovery effect of lower limb hip and knee joints of different patients. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art 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 embodiments can be obtained from these drawings without creative effort.

[0054] Figure 1 A flowchart of the first method for testing lower limb hip and knee training devices based on joint motion coordination is shown;

[0055] Figure 2 A second method flowchart for testing lower limb hip and knee training devices based on joint motion coordination is shown;

[0056] Figure 3 A system framework diagram of a testing system for lower limb hip and knee training equipment based on joint motion coordination is shown. Detailed Implementation

[0057] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0058] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0059] The first aspect of this invention provides a testing method for lower limb hip and knee training devices based on joint movement coordination, such as... Figure 1 As shown, it includes the following steps:

[0060] S102: Based on the probability calculation of the correct movement mechanism in human anatomy, multidimensional injury data of the patient subject is used to constrain the tensor distribution of normal movement of the hip and knee joints of the lower limbs, so as to configure the device for a dedicated training strategy for the patient subject.

[0061] S104: Using a three-dimensional simulation motion model of the lower limb hip and knee device, perform coordinated injury training simulation on the patient, obtain muscle group activation simulation signals, analyze the trend evolution of the muscle group activation simulation signals, and obtain the actual muscle group activation status of the patient.

[0062] S106: Construct an activation state recognition and transfer map by combining different preset muscle group activation states with corresponding activation transfer indices. Index the actual muscle group activation status through the activation state recognition and transfer map to obtain the current movement mode during equipment training.

[0063] S108: Based on the current motion mode, the motion posture is transmitted and adjusted on the standard joint execution chain of the hip and knee of the human lower limb to generate a standard motion coordination trajectory. Based on the standard motion coordination trajectory, the motion coordination of the simulated motion coordination trajectory of the device is judged to determine whether the motion coordination is qualified, and the first test result is obtained.

[0064] S110: Based on the first test result, a specified motion scenario is arranged to perform a second test on the lower limb hip and knee training equipment at the zero torque point and to determine the instantaneous control of gait torque, thereby obtaining the second test result.

[0065] More specifically, step S102 includes the following steps:

[0066] Obtain patient case report information, extract multidimensional injury data of the patient's lower limb hip and knee joints from the case report information, and obtain the correct movement mechanism of the lower limb hip and knee joints based on big data network;

[0067] Based on multidimensional injury data, construct injury tensor matrices with different motion dimensions, calculate the covariance matrix of the injury tensor matrix, and obtain the constraint eigenvalues ​​and constraint eigenvectors of multidimensional injury data along different motion dimensions through the covariance matrix.

[0068] Based on the constraint feature value, the kernel function and bandwidth of the constraint feature vector are preset. The probability density estimate of the injury constraint that causes normal movement of the hip and knee joints of the lower limbs in different motion dimensions is calculated by using the kernel function and bandwidth.

[0069] A singular value decomposition algorithm is introduced. Based on the correct motion mechanism, the constraint feature vector of the injury tensor matrix is ​​decomposed in the singular value decomposition algorithm to obtain the core tensor of the injury degree in different motion dimensions and the factor matrix that the core tensor in different motion dimensions restricts the normal motion of the patient's lower limb hip and knee joints.

[0070] Obtain the patient's exclusive training task, preset the minimum task loss function according to the exclusive training task, update the core tensor and factor matrix based on the decomposition rank alternation constraint of the injury constraint probability density estimate until the minimum task loss function is reached, and generate the tensor distribution of the injury impact on the normal movement of the patient's lower limb hip and knee joints in different movement dimensions when the injury severity of the injury data generates multidimensional injury data.

[0071] The motion structure of the lower limb hip and knee training device is obtained. Based on the motion structure, the lower limb hip and knee joints of the patient are divided into N sub-joint process blocks. Training decision allocation is performed on each sub-joint process block based on the injury influence tensor distribution, and a special training strategy for the lower limb hip and knee training device for the patient is generated.

[0072] It should be noted that multidimensional injury data refers to the extent of injury to the hip and knee joints of the lower limbs in terms of both time and space. The degree of injury to the hip and knee joints varies among patients, but traditional lower limb hip and knee training equipment typically has a relatively simple control mode. This makes it difficult to develop highly tailored training strategies for patients with different injury levels. Consequently, patients with varying injury levels are subjected to the same parameters such as joint flexion and extension angles, flexion and extension rates, or flexion and extension forces during training. This may cause different patients to experience training pain or discomfort. Therefore, it is necessary to tailor highly specific training strategies to the individual patient's injury level. To address this, this method first acquires multidimensional injury data of the patient's lower limb hip and knee joints. Since lower limb hip and knee joint training movements under injury conditions may exhibit a certain probability distribution of localized pain or movement constraints in a specific time and space, the method constructs injury tensor matrices for different movement dimensions using the multidimensional injury data. The covariance matrix of these injury tensors is then calculated. This covariance matrix represents the global distribution structure of the multidimensional injury data constraining lower limb hip and knee joint movement along different movement dimensions. The constraint eigenvalues ​​and constraint eigenvectors recorded in the matrix express the direction of the probability of localized pain or discomfort restricting joint movement during lower limb hip and knee joint movement in both time and space. This provides a reliable pattern inference basis for subsequent analysis of random injury constraints that may lead to normal lower limb hip and knee joint movement in different movement dimensions. Therefore, the kernel function and bandwidth of the constraint eigenvector can be preset based on the constraint eigenvalues ​​to calculate the probability density estimate of injury constraints leading to normal lower limb hip and knee joint movement in different movement dimensions. This improves the device's ability to determine reasonable training control modes for random joint movement constraints that may arise from different patient injuries.

[0073] It should be noted that pain and discomfort associated with hip and knee joint movement can locally transfer to bones or muscle groups. For example, if a patient's injury is mild, the flexion and extension of the joint during running while wearing lower limb hip and knee training equipment may cause slight soreness in a local bone. Therefore, it is necessary to calculate the detailed distribution of the injury's impact based on the severity of the injury to provide highly reliable support for the adaptive training control of the equipment. Thus, this method decomposes the constraint feature vector of the injury tensor matrix using a singular value decomposition algorithm based on the correct movement mechanism. This efficiently obtains a core tensor and factor matrix regarding the constraint of injury severity on normal hip and knee joint movement in different movement dimensions. The core tensor represents the constraint volume of injury severity under a multidimensional injury data structure, while the factor matrix consists of local features of the injury manifestation that restrict normal hip and knee joint movement under this constraint volume. It is a dimension-reduced basis vector, which can better define the injury data in different movement dimensions. Together, they form an expression system for the limitation of local joint movement under different injury degrees. Next, by minimizing the pre-defined task loss function for the specific training task, the core tensor and factor matrix are alternately constrained and updated. This allows for a more accurate representation of the random distribution of injury severity revealed by multidimensional injury data during movement, particularly affecting the hip and knee joints and related areas of the lower limbs. The resulting injury impact tensor distribution represents the local constraint influence of the injury severity in the multidimensional injury data on the normal movement of the patient's hip and knee joints in different movement dimensions, corresponding to the slight soreness experienced in a specific bone area. Finally, the training control decisions for the lower limb hip and knee training device are further allocated based on this injury impact tensor distribution. This method enables the development of comfortable and highly compatible training strategies tailored to patients with different injury severity levels, resulting in more accurate test results when the device is applied to different patients. This provides sufficient test reliability for subsequent optimization of the device's training control for different patients.

[0074] More specifically, step S104 includes the following steps:

[0075] The assembly design drawings of the lower limb hip and knee training equipment and the joint injury movement process after the patient's multidimensional injury data were obtained. Based on the assembly design drawings, a three-dimensional simulation motion model of the lower limb hip and knee equipment was constructed in SolidWorks modeling software. Based on the joint injury movement process, a joint coordination motion model of the lower limb hip and knee joint injury was constructed.

[0076] The joint coordination motion model of the patient's lower limb hip and knee joint after injury is simulated and tested by controlling the three-dimensional simulation running model to execute the exclusive training strategy. During the simulation test, electromyography technology is used to record the muscle activation status in sequence to obtain the simulated activation signals of multiple muscle groups of the patient.

[0077] Sine waveform spectra of each group of muscle group activation simulation signals are extracted. Based on the sinusoidal waveform spectra, a Hanning window function that balances the time-frequency resolution of muscle group activation is preset. The instantaneous autocorrelation function of the signal transition is calculated through the complex conjugate of the muscle group activation simulation signal. Based on the Hanning window function, the time delay parameter of the instantaneous autocorrelation function is transformed to obtain the time-frequency simulation distribution characterization of multiple groups of muscle group activation simulation signals output by the hip and knee joints of the patient's lower limbs.

[0078] The local maximum peak point and local minimum peak point of each muscle group activation simulation signal are detected and obtained. The upper and lower envelopes of the muscle group activation simulation signal are constructed by spline interpolation of the local maximum peak point and local minimum peak point. At this time, the balance value between the upper and lower envelopes is calculated.

[0079] A regeneration signal is output by removing the equalization value from multiple muscle group activation simulation signals. If the upper envelope of the regeneration signal is completely symmetrical with respect to the lower envelope, the intrinsic modes are continuously extracted from the signal according to the time-frequency simulation distribution characterization to form a time-frequency distribution intrinsic mode chain. The actual muscle group activation status of the patient during the lower limb hip and knee training device training is determined according to the time-frequency distribution intrinsic mode chain.

[0080] It should be noted that there is a close relationship between joint movement and muscle activation. The muscles associated with the hip and knee generate force through contraction, driving the hip and knee joints to perform flexion, extension, or rotation movements. Therefore, analyzing the activation status of muscle groups can indirectly reflect the training quality of lower limb hip and knee training equipment for patients, thus further clarifying the testing performance of the equipment. To this end, this method uses three-dimensional simulation to test the joint coordination training of lower limb hip and knee training equipment under the premise of a patient injury. This replaces the traditional multi-step on-site sampling test process, reducing significant manpower and labor costs. Furthermore, it reduces the intervention error of manual testing, effectively improving the efficiency and flexibility of test data acquisition, ensuring accurate and reliable test results. Next, the time-frequency distribution characteristics of the dispersed force generated by muscle activation are extracted from the activation signals of the lower limb hip and knee joints recorded during the simulation. Simultaneously, spline interpolation is performed based on the local maximum and minimum peak points of each group of muscle activation simulation signals to construct the upper and lower envelopes of the muscle activation simulation signals, making the local features of the signal clearer and thus determining the oscillation range of the signal. The equilibrium value between these two factors measures the local transmission trend of the signal, which is a crucial prerequisite for highlighting the muscle activation trend. By subtracting this equilibrium value from multiple sets of simulated muscle activation signals, a narrow-band oscillation mode with physical significance guiding muscle activation can be isolated. This mode has a more singular frequency component and can better reflect the development of joint movement driven by muscle activation during the test. Furthermore, analyzing the local equilibrium of the signal through the symmetry of the envelope helps to separate the time-frequency distribution components of the signal at different scales, thereby converging them into a complete time-frequency distribution intrinsic mode chain of the muscle activation signal trend. This significantly improves the accuracy of inferring the actual muscle activation state from signal data.

[0081] It should be noted that although traditional feature extraction of signal time-frequency distribution has good time-frequency resolution, it is prone to generating transitive cross-terms. To address this, our method introduces a Hanning window function that balances the time-frequency resolution of muscle activation to suppress cross-terms, maintaining relatively clear time-frequency resolution and reducing unnecessary transitive cross-term interference. The instantaneous autocorrelation function represents the autocorrelation characteristics of a signal at a specific time point, revealing the joint similarity of the signal under different time delays. The accuracy of this instantaneous autocorrelation function is crucial, directly affecting the accuracy of the time-frequency distribution. In summary, this method can simulate the training and testing of a patient's injured lower limb hip and knee joint using a lower limb hip and knee training device, and perform situational evolution inference on the output muscle activation signals. It can characterize the coordinated joint movement during the test based on the muscle activation status, thus providing analytical clues for the training control qualification of the lower limb hip and knee training device and significantly improving the testing accuracy of the device.

[0082] More specifically, step S106 includes the following steps:

[0083] Based on big data networks, we obtain ergonomic knowledge graphs and muscle group movement cases of the hip and knee joints of the lower limbs. We then extract several movement patterns when different preset muscle group activation states occur in the hip and knee joints of the lower limbs through the ergonomic knowledge graphs.

[0084] By extracting the historical movement trend span of each preset muscle group activation state that causes the lower limb hip and knee joint movements to produce a coordinated pattern through muscle group movement cases, activation transfer index of different preset muscle group activation states is preset based on the historical movement trend span.

[0085] Each preset muscle group activation state is defined as a state node, and the activation transition index is used as the index boundary between each state node and the corresponding movement pattern. An activation state recognition transition map is constructed by fitting the state nodes and the index boundary.

[0086] A trend change detection model is constructed, and the actual muscle group activation status of the patient during the training of the lower limb hip and knee training device is accumulated and calculated using the trend change detection model to obtain the current activation status accumulation sum.

[0087] If the current active state cumulative sum is detected to be greater than the preset cumulative threshold, then the trend change length of the current active state cumulative sum is obtained. Based on the current active state cumulative sum, the trend change length is weighted and assigned a smooth character to obtain the active state string about the state transition at different trend change nodes.

[0088] Based on the activation transition index, the next state of each character in the activation state string is identified by traversing the activation trend recognition transition map. The acceptance coefficient of the activation trend recognition transition map for each character is output. If the acceptance coefficient is greater than the preset acceptance coefficient, the current movement mode of the patient when using the lower limb hip and knee training device is determined according to the index boundary of the current character state.

[0089] It is important to note that different muscles work together during joint movement to achieve smooth and stable joint motion. However, if joint mobility is restricted, muscle activation may become unbalanced, leading to compensatory movements and even increasing the risk of injury. Therefore, different movement patterns activate different muscle groups. The characteristics of muscle activation in different movement patterns are significant; for example, leg extension primarily activates the quadriceps. Therefore, for joint training and recovery, it is crucial to identify the movement patterns that drive the response of lower limb hip and knee training equipment. To address this, this method uses an activation state recognition transition map to identify patterns in the deduced actual muscle group activation states. Since different movement patterns generate different preset muscle group activation states, and there is a predetermined connection index between the two, this paper extracts the historical movement trend span of each preset muscle group activation state, which leads to the coordinated movement of the hip and knee joints of the lower limbs. This historical movement trend span is the degree of trend jump from one muscle group activation state to another, showing the coordinated and non-coordinated movement patterns of the hip and knee joints. It can clearly reflect the activation transfer amplitude of muscle group activation states under the definition of different movement patterns. Therefore, activation transfer indices for different preset muscle group activation states can be preset based on the historical movement trend span, providing a joint transformation label for the movement pattern recognition of muscle group activation transfer. That is, the activation transfer index is used as the index boundary between each state node and the corresponding movement pattern. The activation state recognition transfer map can intuitively show the index relationship between the changes in muscle group activation transfer states and movement patterns. Furthermore, the index relationship between the two is stored virtually as a character, which can accurately identify the corresponding movement pattern index under different muscle group activation trends, optimize the character index efficiency of muscle group activation states, and improve the membership matching accuracy of movement patterns related to muscle group activation states.

[0090] It should be noted that since the activation state recognition transition map is performed using character recognition, the recognition of the actual muscle group activation state should also be performed in character form. Therefore, the actual muscle group activation state needs to be characterized. To address this, this method constructs a trend change detection model to cumulatively calculate the trend transition rules of the actual muscle group activation state, obtaining the current activation state cumulative sum. If the current activation state cumulative sum is greater than a preset cumulative threshold, it indicates that there is a significant change in the state transition of the currently accumulated muscle group activation state. This requires a specific character length definition. Therefore, the trend change length is smoothed by weighted assignment of characters to generate activation state strings that indicate the state transition of the activation state at different trend change nodes. Characterization can adapt to the activation state recognition transition map for the recognition of the actual muscle group activation state, and can also improve recognition efficiency. The string is a unique label for different state transitions, achieving targeted recognition of the movement patterns corresponding to the muscle group activation state. Finally, the next state of each character in the weighted activation state string is identified through an activation state recognition transition map. If the acceptance coefficient is greater than the preset acceptance coefficient, it indicates that the index boundary has a high acceptance rate for that state and is in an accepting state. Therefore, the movement pattern corresponding to the index boundary is the identification result that highly belongs to the actual muscle group activation state. This method can identify the corresponding movement patterns for the activation trend of muscle groups during the testing process, thereby facilitating the analysis of the lower limb hip and knee training equipment's movement coordination of the lower joints and improving the accuracy of equipment training tests.

[0091] More specifically, step S108, as follows: Figure 2 As shown, the specific steps include:

[0092] S202: Obtain standard joint structure diagrams and joint activity mechanisms of the human lower limb hip and knee based on big data networks, and construct standard joint execution chains of the human lower limb hip and knee based on the activities of the standard joint structure diagrams on the execution constraints, using the joint activity mechanisms as execution constraints.

[0093] S204: Obtain the injury characteristics of the patient's lower limb hip and knee joints based on multidimensional injury data, import the injury characteristics and the current movement mode into the ergonomic knowledge graph for identification, and output the maximum extension landing point of the patient's maximum movement under the current movement mode and the free movement width between adjacent joint connection structures in the standard joint structure diagram.

[0094] S206: Remove the first and last joints that are closest to the limit of extension on the standard joint execution chain, construct the root node of the first joint and the end effector of the end joint, and move the end effector to the limit of extension starting from the root node, based on the preset limit of the transmission length threshold of the free range of motion.

[0095] S208: If the current transmission length between adjacent joint connection structures exceeds the limited transmission length threshold, adjust the motion pose between adjacent joint connection structures until the current transmission length between adjacent joint connection structures does not exceed the limited transmission length threshold.

[0096] S210: Repeat the above steps to generate a standard motion coordination trajectory of the patient in the current motion mode under the condition of lower limb hip and knee joint injury. Based on the standard motion coordination trajectory, determine the motion rhythm and coordination of the lower limb hip and knee training device on the simulated motion coordination trajectory to obtain the first test result.

[0097] It should be noted that identifying the patient's movement pattern during simulated training allows for the clarification of the movement patterns of the hip and knee joints in the lower limbs. Under this premise, the hip and knee joint movements maintain a maximum reasonable trajectory. When the simulated trajectory deviates significantly from this reasonable trajectory, it indicates a serious control error in the coordination and rhythm of the joint synergistic movements driven by the lower limb hip and knee training equipment, thus rendering the equipment unqualified. Therefore, this method first constructs a standard joint execution chain for the human lower limb hip and knee. This standard joint execution chain follows the regular movement execution of the standard joint structure of the human lower limb hip and knee under the joint activity mechanism specifications, and is a joint activity chain highly consistent with human anatomy, providing a standardized definition basis for the subsequent calculation of the standard synergistic movement trajectory. Next, based on the patient's injury characteristics and current movement pattern, the system identifies the maximum extension point of the patient's movement within that pattern, as well as the free range of motion between adjacent joints in a standard joint anatomy diagram. The maximum extension point represents the maximum target position where the patient can extend their joints to their maximum extent under the injury severity conditions; while the free range of motion represents the range of extension lengths between joints. Both are patient-centered training trajectory reference and convergence standards, avoiding infinite loops and enabling the creation of a dedicated test reference comparison group for patients with different injury degrees, significantly improving the test comparison diversity of lower limb hip and knee training equipment.

[0098] It should be noted that by fixing the root node and first moving the end effector to its maximum extension point, if the current transmission length between adjacent joint connections exceeds the defined transmission length threshold, it indicates that the joint pose of the current standard joint execution chain does not conform to the coordinated movement of the flexion and extension range of motion of the bones under the patient's injury state. Therefore, each joint needs to be adjusted sequentially, either increasing or decreasing, to maintain a fixed bone length constraint and gradually move closer to the root. This allows the end effector to get closer to the target while maintaining the distance constraint between joints, ensuring that the root node of the bones does not drift, and significantly improving the accuracy and coordination of the standard motion coordination trajectory planning. This method can construct a standard joint execution chain for the hip and knee of the human lower limb and redistribute joint positions according to the injury characteristics and movement patterns of different patients. This allows them to meet the length constraint while bringing the end effector of the bones closer to the maximum position of maximum extension, thereby making the test comparison reference of the lower limb hip and knee training equipment more accurate and reliable, and significantly improving the stability and accuracy of the test of the coordination and rhythm of the lower limb hip and knee training equipment.

[0099] More specifically, the process of repeating the above steps to generate a standard motion coordination trajectory of the patient's lower limb hip and knee joint under the current motion mode, and then judging the motion rhythm and coordination of the lower limb hip and knee training device based on the standard motion coordination trajectory to obtain the first test result, specifically includes the following steps:

[0100] The arrival error tolerance of the preset limit extension landing point is repeated, and the above steps of moving the end effector to the limit extension landing point and adjusting the motion posture between adjacent joint connection structures are repeated to obtain the current arrival distance of the end effector relative to the limit extension landing point.

[0101] If the current arrival distance is less than the arrival error tolerance, stop the motion posture adjustment operation, generate the standard motion coordination trajectory of the patient in the current motion mode under the condition of lower limb hip and knee joint injury, and obtain the simulated motion coordination trajectory of the patient during the training process of the lower limb hip and knee training device.

[0102] By extracting key collaborative nodes of the lower limb knee joint following the joint activity mechanism from the ergonomic knowledge graph, the entropy weight algorithm is introduced to calculate the approximate entropy of each key collaborative node on the simulated motion collaborative trajectory compared to the standard motion collaborative trajectory, resulting in multiple sparse approximate entropies and multiple dense approximate entropies.

[0103] Calculate the ratio of the number of key collaborative nodes between sparse approximation entropy and dense approximation entropy, and determine whether the ratio is greater than a preset ratio.

[0104] If the quantity ratio is less than the preset quantity ratio, the lower limb hip and knee training equipment is calibrated as qualified equipment; if the quantity ratio is greater than the preset quantity ratio, the lower limb hip and knee training equipment is calibrated as unqualified equipment, and the first test result is obtained.

[0105] It should be noted that, for the generation of the standard motion coordination trajectory, this method presets an arrival error tolerance for the ultimate extension landing point. This arrival error tolerance determines the convergence standard for the adjustment of each joint in the standard joint execution chain, avoiding infinite adjustment loops and reducing the deviation of the standard motion coordination trajectory. If the current arrival distance is less than the arrival error tolerance, it means that the last joint of the standard joint execution chain after joint adjustment has gradually approached the maximum target position of ultimate extension. This indicates that the adjusted joint execution chain conforms to the standardized motion trajectory of the joint in the current motion mode under the actual injury state of the patient, and therefore no further adjustment is needed. As for the test deviation judgment of the simulated motion coordination trajectory compared with the standard motion coordination trajectory, this method calculates the approximate entropy of each key coordination node on the simulated motion coordination trajectory compared with the standard motion coordination trajectory. The key coordination node refers to the node when the hip and knee joints of the lower limb have highly coordinated movement. This node can be a muscle connection node or a joint splicing point on the joint. In short, this key coordination node reflects the key layout of the lower limb hip and knee joint motion coordination. The calculated sparse and dense approximate entropies reflect the degree of deviation of the simulated motion coordination trajectory from the standard motion coordination trajectory. Sparse approximate entropy indicates that the coordinate points of the simulated motion coordination trajectory are sparser compared to the standard trajectory, while dense approximate entropy indicates that the coordinate points of the relative deviation are denser. If the ratio is less than the preset ratio, it indicates that the lower limb hip and knee training device provides high coordination and stable rhythm in the patient's joint movements during training, thus the device is qualified. Conversely, if the ratio is greater than the preset ratio, it indicates that the lower limb hip and knee training device results in poor joint coordination and rhythm during running or walking, extremely unnatural movements, difficulty in achieving training goals, and a potential aggravation of the patient's injury, thus the device is unqualified. This method allows for comparative analysis of the simulated trajectory during testing against the standard motion coordination trajectory under the patient's injury state, thereby efficiently and quickly determining whether the coordination and rhythm of the lower limb hip and knee training device during patient training are qualified, improving the accuracy and reliability of quality screening for lower limb hip and knee training devices.

[0106] More specifically, step S110 includes the following steps:

[0107] If the first test result shows that the lower limb hip and knee training equipment is qualified, then a specified motion scenario is set up in the simulation process to test the lower limb hip and knee training equipment and obtain multiple sets of joint force sensing test parameters of the lower limb hip and knee training equipment.

[0108] By introducing Newton's laws of dynamics to construct a dynamic topology model, and using multiple sets of joint force sensing test parameters, the reaction force between the lower limb hip-knee movement gait and the specified movement scenario is calculated within the dynamic topology model. This yields the simulated zero-moment point trajectory of the patient's movement gait under the action of the lower limb hip and knee training device during the simulation process.

[0109] Obtain the expected test requirements of the lower limb hip and knee training device and the drive control base station, and plan the expected zero-moment point trajectory of the lower limb hip and knee training device to apply to the patient under the premise of a specified motion scenario based on the expected test requirements;

[0110] Based on the preset gait torque stations of the drive control base station, the deviation between the simulated zero torque point trajectory and the expected zero torque point trajectory on each gait torque station is calculated, and multiple misalignment deviation values ​​are obtained. Only the gait torque stations with misalignment deviation values ​​greater than the preset misalignment deviation threshold are extracted and defined as suspicious gait torque stations.

[0111] Construct a transient peak fluctuation domain, fit suspicious gait torque sites within the transient peak fluctuation domain, and output the actual transient peak fluctuation tomography triggered by suspicious gait torque during the simulation of the lower limb hip and knee training device;

[0112] If the actual instantaneous peak fluctuation fault is at least one level higher than the preset instantaneous peak fluctuation fault, it indicates that the instantaneous torque control performance of the lower limb hip and knee training device is poor, resulting in gait coordination error, and the second test result is obtained.

[0113] It should be noted that the joint force sensing test parameters include joint angle, velocity, acceleration, or plantar force. When the lower limb hip and knee training device is qualified in the initial test, it indicates that the movement coordination and rhythm changes during training have a certain degree of stability. Therefore, in order to explore the accuracy of its control in depth, it is necessary to test and analyze its instantaneous control in maintaining the patient's training gait. To this end, this method tests the change of the zero torque point of the lower limb hip and knee training device driving the patient's hip and knee joint movement by applying a specific motion scenario. The zero torque point refers to a specific point within the support surface where the torque generated by the ground reaction force about the horizontal coordinate axis is zero. Simply put, when an object (such as a robot or human body) stands or walks on the ground, the ground will generate a reaction force on the object. This reaction force can be decomposed into a vertical force and a horizontal force. There is a point within the support surface where the rotational effect (i.e., torque) of the ground reaction force about the horizontal direction is zero. When lower limb hip and knee training equipment moves a patient, it generates a zero-torque point. Changes in this zero-torque point are closely related to changes in gait. If the simulated zero-torque point deviates significantly from the expected zero-torque point trajectory, it indicates an abnormality in the instantaneous torque control of the lower limb hip and knee training equipment, leading to extreme gait instability or incorrect angle parameters. For the analysis of this deviation, this method uses a pre-set gait torque station based on the drive control base station. The drive control base station comprises the drive motor, transmission device, and other components on the lower limb hip and knee training equipment that implement torque control; this serves as the control benchmark for determining the rationality of the gait movement driven by the equipment. Since torque has instantaneous control characteristics, this method constructs an instantaneous peak fluctuation domain to fit the zero-torque point control expression of the suspected gait torque station. The output instantaneous peak fluctuation tomography reflects the hierarchical morphology of the peak fluctuations generated by the simulated zero-torque point trajectory due to gait torque during the simulation, providing a comprehensive and detailed representation of the actual instantaneous torque control performance of the lower limb hip and knee training equipment. This method can test and evaluate the gait torque control performance of lower limb hip and knee training equipment, thereby achieving instantaneous control qualification screening of lower limb hip and knee training equipment and ensuring the evaluation and optimization of equipment control quality.

[0114] The second aspect of this invention provides a testing system for lower limb hip and knee training devices based on joint motion coordination, such as... Figure 3 As shown, the lower limb hip and knee training device testing system includes a memory 31 and a processor 32. The memory 31 stores a lower limb hip and knee training device testing method program based on joint movement coordination. When the lower limb hip and knee training device testing method program is executed by the processor 32, any of the steps of the lower limb hip and knee training device testing method described above are implemented.

[0115] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A testing method for lower limb hip and knee training devices based on joint motion coordination, characterized in that, Includes the following steps: S102: Based on the probability calculation of the correct movement mechanism in human anatomy, multidimensional injury data of the patient subject is used to constrain the tensor distribution of normal movement of the hip and knee joints of the lower limbs, so as to configure the device for a dedicated training strategy for the patient subject. S104: Using a three-dimensional simulation motion model of the lower limb hip and knee device, perform coordinated injury training simulation on the patient, obtain muscle group activation simulation signals, analyze the trend evolution of the muscle group activation simulation signals, and obtain the actual muscle group activation status of the patient. S106: Construct an activation state recognition and transfer map by combining different preset muscle group activation states with corresponding activation transfer indices. Index the actual muscle group activation status through the activation state recognition and transfer map to obtain the current movement mode during equipment training. S108: Based on the current motion mode, the motion posture is transmitted and adjusted on the standard joint execution chain of the hip and knee of the human lower limb to generate a standard motion coordination trajectory. Based on the standard motion coordination trajectory, the motion coordination of the simulated motion coordination trajectory of the device is judged to determine whether the motion coordination is qualified, and the first test result is obtained. S110: Based on the first test result, a specified motion scenario is arranged to perform a second zero-torque point test on the lower limb hip and knee training equipment and determine the instantaneous control of gait torque to obtain the second test result; Specifically, step S106 includes the following steps: Based on big data networks, we obtain ergonomic knowledge graphs and muscle group movement cases of the hip and knee joints of the lower limbs. We then extract several movement patterns when different preset muscle group activation states occur in the hip and knee joints of the lower limbs through the ergonomic knowledge graphs. By extracting the historical movement trend span of each preset muscle group activation state that causes the lower limb hip and knee joint movements to produce a coordinated pattern through muscle group movement cases, activation transfer index of different preset muscle group activation states is preset based on the historical movement trend span. Each preset muscle group activation state is defined as a state node, and the activation transition index is used as the index boundary between each state node and the corresponding movement pattern. An activation state recognition transition map is constructed by fitting the state nodes and the index boundary. A trend change detection model is constructed, and the actual muscle group activation status of the patient during the training of the lower limb hip and knee training device is accumulated and calculated using the trend change detection model to obtain the current activation status accumulation sum. If the current active state cumulative sum is detected to be greater than the preset cumulative threshold, then the trend change length of the current active state cumulative sum is obtained. Based on the current active state cumulative sum, the trend change length is weighted and assigned a smooth character to obtain the active state string about the state transition at different trend change nodes. Based on the activation transition index, the next state of each character in the activation state string is identified by traversing the activation trend recognition transition map. The acceptance coefficient of the activation trend recognition transition map for each character is output. If the acceptance coefficient is greater than the preset acceptance coefficient, the current movement mode of the patient when using the lower limb hip and knee training device is determined according to the index boundary of the current character state.

2. The testing method for the lower limb hip and knee training device based on joint motion coordination according to claim 1, characterized in that, Step S102 specifically includes the following steps: Obtain patient case report information, extract multidimensional injury data of the patient's lower limb hip and knee joints from the case report information, and obtain the correct movement mechanism of the lower limb hip and knee joints based on big data network; Based on multidimensional injury data, construct injury tensor matrices with different motion dimensions, calculate the covariance matrix of the injury tensor matrix, and obtain the constraint eigenvalues ​​and constraint eigenvectors of multidimensional injury data along different motion dimensions through the covariance matrix. Based on the constraint feature value, the kernel function and bandwidth of the constraint feature vector are preset. The probability density estimate of the injury constraint that causes normal movement of the hip and knee joints of the lower limbs in different motion dimensions is calculated by using the kernel function and bandwidth. A singular value decomposition algorithm is introduced. Based on the correct motion mechanism, the constraint feature vector of the injury tensor matrix is ​​decomposed in the singular value decomposition algorithm to obtain the core tensor of the injury degree in different motion dimensions and the factor matrix that the core tensor in different motion dimensions restricts the normal motion of the patient's lower limb hip and knee joints. Obtain the patient's exclusive training task, preset the minimum task loss function according to the exclusive training task, update the core tensor and factor matrix based on the decomposition rank alternation constraint of the injury constraint probability density estimate until the minimum task loss function is reached, and generate the tensor distribution of the injury impact on the normal movement of the patient's lower limb hip and knee joints in different movement dimensions when the injury severity of the injury data generates multidimensional injury data. The motion structure of the lower limb hip and knee training device is obtained. Based on the motion structure, the lower limb hip and knee joints of the patient are divided into N sub-joint process blocks. Training decision allocation is performed on each sub-joint process block based on the injury influence tensor distribution, and a special training strategy for the lower limb hip and knee training device for the patient is generated.

3. The testing method for the lower limb hip and knee training device based on joint motion coordination according to claim 1, characterized in that, Step S104 specifically includes the following steps: The assembly design drawings of the lower limb hip and knee training equipment and the joint injury movement process after the patient's multidimensional injury data were obtained. Based on the assembly design drawings, a three-dimensional simulation motion model of the lower limb hip and knee equipment was constructed in SolidWorks modeling software. Based on the joint injury movement process, a joint coordination motion model of the lower limb hip and knee joint injury was constructed. The joint coordination motion model of the patient's lower limb hip and knee joint after injury is simulated and tested by controlling the three-dimensional simulation running model to execute the exclusive training strategy. During the simulation test, electromyography technology is used to record the muscle activation status in sequence to obtain the simulated activation signals of multiple muscle groups of the patient. Sine waveform spectra of each group of muscle group activation simulation signals are extracted. Based on the sinusoidal waveform spectra, a Hanning window function that balances the time-frequency resolution of muscle group activation is preset. The instantaneous autocorrelation function of the signal transition is calculated through the complex conjugate of the muscle group activation simulation signal. Based on the Hanning window function, the time delay parameter of the instantaneous autocorrelation function is transformed to obtain the time-frequency simulation distribution characterization of multiple groups of muscle group activation simulation signals output by the hip and knee joints of the patient's lower limbs. The local maximum peak point and local minimum peak point of each muscle group activation simulation signal are detected and obtained. The upper and lower envelopes of the muscle group activation simulation signal are constructed by spline interpolation of the local maximum peak point and local minimum peak point. At this time, the balance value between the upper and lower envelopes is calculated. A regeneration signal is output by removing the equalization value from multiple muscle group activation simulation signals. If the upper envelope of the regeneration signal is completely symmetrical with respect to the lower envelope, the intrinsic modes are continuously extracted from the signal according to the time-frequency simulation distribution characterization to form a time-frequency distribution intrinsic mode chain. The actual muscle group activation status of the patient during the lower limb hip and knee training device training is determined according to the time-frequency distribution intrinsic mode chain.

4. The testing method for the lower limb hip and knee training device based on joint motion coordination according to claim 1, characterized in that, Step S108 specifically includes the following steps: Based on big data networks, standard joint structure diagrams and joint activity mechanisms of the human lower limb hip and knee are obtained. Using the joint activity mechanisms as execution constraints, standard joint execution chains of the human lower limb hip and knee are constructed based on the activities of the standard joint structure diagrams on the execution constraints. Based on multidimensional injury data, the injury characteristics of the patient's lower limb hip and knee joints are obtained. The injury characteristics and the current movement pattern are imported into the ergonomic knowledge graph for identification. The maximum extension landing point of the patient's maximum movement under the current movement pattern and the free movement width between adjacent joint connection structures in the standard joint structure diagram are output. The first and last joints closest to the limit of extension on the standard joint execution chain are separated, and the root node of the first joint and the end effector of the end joint are constructed. Based on the preset limit of the transmission length threshold of the free range of motion, the end effector is moved to the limit of extension starting from the root node. If the current transmission length between adjacent joint connection structures exceeds the limit transmission length threshold, the motion pose between adjacent joint connection structures is adjusted until the current transmission length between adjacent joint connection structures does not exceed the limit transmission length threshold. Repeat the above steps to generate a standard motion coordination trajectory of the patient in the current motion mode under the condition of lower limb hip and knee joint injury. Based on the standard motion coordination trajectory, determine the motion rhythm and coordination of the lower limb hip and knee training device on the simulated motion coordination trajectory to obtain the first test result.

5. The testing method for the lower limb hip and knee training device based on joint motion coordination according to claim 4, characterized in that, The process involves repeating the above steps to generate a standard motion coordination trajectory for the patient's lower limb hip and knee joint injury in the current movement mode. Based on this standard motion coordination trajectory, the motion rhythm and coordination of the lower limb hip and knee training device are determined using the simulated motion coordination trajectory to obtain the first test result. This process specifically includes the following steps: The arrival error tolerance of the preset limit extension landing point is repeated, and the above steps of moving the end effector to the limit extension landing point and adjusting the motion posture between adjacent joint connection structures are repeated to obtain the current arrival distance of the end effector relative to the limit extension landing point. If the current arrival distance is less than the arrival error tolerance, stop the motion posture adjustment operation, generate the standard motion coordination trajectory of the patient in the current motion mode under the condition of lower limb hip and knee joint injury, and obtain the simulated motion coordination trajectory of the patient during the training process of the lower limb hip and knee training device. By extracting key collaborative nodes of the lower limb knee joint following the joint activity mechanism from the ergonomic knowledge graph, the entropy weight algorithm is introduced to calculate the approximate entropy of each key collaborative node on the simulated motion collaborative trajectory compared to the standard motion collaborative trajectory, resulting in multiple sparse approximate entropies and multiple dense approximate entropies. Calculate the ratio of the number of key collaborative nodes between sparse approximation entropy and dense approximation entropy, and determine whether the ratio is greater than a preset ratio. If the quantity ratio is less than the preset quantity ratio, the lower limb hip and knee training equipment is calibrated as qualified equipment; if the quantity ratio is greater than the preset quantity ratio, the lower limb hip and knee training equipment is calibrated as unqualified equipment, and the first test result is obtained.

6. The testing method for the lower limb hip and knee training device based on joint motion coordination according to claim 1, characterized in that, Step S110 specifically includes the following steps: If the first test result shows that the lower limb hip and knee training equipment is qualified, then a specified motion scenario is set up in the simulation process to test the lower limb hip and knee training equipment and obtain multiple sets of joint force sensing test parameters of the lower limb hip and knee training equipment. By introducing Newton's laws of dynamics to construct a dynamic topology model, and using multiple sets of joint force sensing test parameters, the reaction force between the lower limb hip-knee movement gait and the specified movement scenario is calculated within the dynamic topology model. This yields the simulated zero-moment point trajectory of the patient's movement gait under the action of the lower limb hip and knee training device during the simulation process. Obtain the expected test requirements of the lower limb hip and knee training device and the drive control base station, and plan the expected zero-moment point trajectory of the lower limb hip and knee training device to apply to the patient under the premise of a specified motion scenario based on the expected test requirements; Based on the preset gait torque stations of the drive control base station, the deviation between the simulated zero torque point trajectory and the expected zero torque point trajectory on each gait torque station is calculated, and multiple misalignment deviation values ​​are obtained. Only the gait torque stations with misalignment deviation values ​​greater than the preset misalignment deviation threshold are extracted and defined as suspicious gait torque stations. Construct a transient peak fluctuation domain, fit suspicious gait torque sites within the transient peak fluctuation domain, and output the actual transient peak fluctuation tomography triggered by suspicious gait torque during the simulation of the lower limb hip and knee training device; If the actual instantaneous peak fluctuation fault is at least one level higher than the preset instantaneous peak fluctuation fault, it indicates that the instantaneous torque control performance of the lower limb hip and knee training device is poor, resulting in gait coordination error, and the second test result is obtained.

7. A testing system for lower limb hip and knee training equipment based on joint motion coordination, characterized in that, The lower limb hip and knee training device testing system includes a memory and a processor. The memory stores a lower limb hip and knee training device testing method program based on joint motion coordination. When the lower limb hip and knee training device testing method program is executed by the processor, the lower limb hip and knee training device testing method steps as described in any one of claims 1-6 are implemented.

Images