Fitness assisting method and power assisting device

Abstract

The invention provides a fitness assisting method and a power assisting device, the power assisting device comprises a controller and an executing mechanism, the method is applied to the controller, and the method comprises the steps that arm force control data of a user is obtained, the arm force control data is analyzed based on a pre-established reflection force arm model, and a reflection force arm model is obtained; obtaining a fitness action type and a training demand of the user; according to the fitness action type and the training requirement, a corresponding adjustment strategy is obtained from a preset power assisting posture adjustment strategy library, and the target posture and angle of a power assisting device are determined; acquiring real-time attitude and angle information of the power assisting device, comparing the real-time attitude and angle information with the target attitude and angle, and calculating deviation values of the attitude and the angle; and whether the deviation value exceeds a preset threshold value or not is judged, and if the deviation value exceeds the preset threshold value, the executing mechanism is adjusted. The method monitors the arm power manipulation change of the user in real time, adjusts the posture of the user in time, and achieves the effect of correcting the fitness posture and angle of the user.

Description

Technical Field

[0001] The present invention relates to the technical field of fitness equipment, and in particular to a fitness auxiliary method and a power-assisting device. Background Art

[0002] During fitness training, the power-assist device needs to adjust its posture and angle in real time according to the user's fitness movements and training needs to provide appropriate assistance. However, the user's fitness movements are dynamically changing, and different movements have different requirements for the posture and angle of the power-assist device. How the power-assist device can accurately sense the user's current fitness movements, and adjust its posture and angle in real time according to the changes in the movements to match the user's fitness movements and training needs, and prevent the user's arm fatigue and ensure user safety, is a key issue in fitness training. Summary of the invention

[0003] The present invention provides a fitness assistance method, wherein the power assisting device comprises a controller and an actuator, and the method is applied to the controller, and the method comprises: Acquire the user's arm force manipulation data, and analyze the arm force manipulation data based on a pre-established reflex arm model to obtain the user's fitness action type and training needs; wherein the arm force manipulation data includes the joint angle and torque of the actuator; According to the fitness action type and training requirements, a corresponding adjustment strategy is obtained from a preset power-assisting posture adjustment strategy library to determine a target posture and angle of the power-assisting device; Acquire the real-time posture and angle information of the power assist device, compare it with the target posture and angle, and calculate the deviation value of the posture and angle; Determine whether the deviation value exceeds a preset threshold value, and if the deviation value exceeds the preset threshold value, adjust the actuator, wherein the deviation value is the adjustment amount of the actuator.

[0004] The technical solution provided by the embodiment of the present invention may have the following beneficial effects: The present invention discloses a fitness assistance method and a power-assisting device. The method monitors the arm force manipulation changes of the user in real time. On the one hand, the power-assisting device assists the user to complete some difficult actions, or increases the resistance through the power-assisting device to achieve the effect of increasing the training intensity. On the other hand, the posture and angle can be evaluated. If the posture and angle of the user deviate, the user posture is adjusted in time to achieve the effect of correcting the user's fitness posture and angle. The present invention realizes intelligent interaction between the fitness power-assisting device and the user, provides personalized and precise posture adjustment, and effectively improves the user's fitness experience and training effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0005] Figure 1 is a flow chart of the fitness assistance method of the present invention; Figure 2 It is a schematic diagram of the actuator of the present invention.

[0006] Reference numerals Motor 1, first connecting arm 2, second connecting arm 3, first connecting member 11, second connecting member 21, third connecting member 31. DETAILED DESCRIPTION

[0007] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0008] like Figure 1-2 The fitness assistance method of this embodiment may specifically include: Step S101, according to the pre-established reflex arm model, the user's current arm force manipulation data is obtained, and the user's current fitness action type and training requirements are obtained by analyzing the arm force manipulation data, and the analysis results are used as the basis for assisting posture adjustment. Among them, the reflex arm model includes parameters such as arm length, mass distribution, and joint friction coefficient. The user's arm force manipulation data is calculated through a dynamic equation, and the arm force manipulation data includes the joint angle and torque of the actuator.

[0009] According to the pre-established reflex arm model, the user's current arm force manipulation data is obtained, including parameters such as arm length, mass distribution, and joint friction coefficient. The user's arm force manipulation data is calculated using the dynamic equation to obtain the user's current arm force manipulation state. The user's current fitness action type is determined by feature extraction and pattern recognition of the arm force manipulation data. According to the user's fitness action type and arm force manipulation state, the user's current training needs and goals are determined. A machine learning algorithm, such as a support vector machine or a decision tree, is used to establish a power-assisted posture adjustment model, which takes the fitness action type and training needs as input and outputs the optimal power-assisted posture parameters. According to the output of the power-assisted posture adjustment model, the exoskeleton or power-assisted device is controlled to adjust the power-assisted posture in real time to provide users with personalized fitness assistance. The user's arm force manipulation data and fitness actions are continuously monitored, and the power-assisted posture adjustment model is dynamically optimized to improve the accuracy and real-time performance of fitness assistance.

[0010] Specifically, first, according to the pre-established reflex arm model, the user's current arm force manipulation data is obtained through the sensor. For example, the length of the user's arm is measured by an infrared sensor to be 65 cm, the mass distribution is 7% of the body weight for the upper arm, 7% for the forearm, and 6% for the hand, and the friction coefficients of the elbow and wrist joints are 0.8 and 0.6 respectively. Then, these parameters are substituted into the dynamic equation, such as the Lagrange equation, to calculate the user's current arm force manipulation state, including joint angle, torque, angular velocity, and angular acceleration. Next, feature extraction is performed on the arm force manipulation data, such as extracting the mean, variance, peak value and other features of the joint angle, and pattern recognition is performed using a support vector machine to determine the user's current fitness action type, such as biceps curl, triceps flexion and extension, etc. According to the identified fitness action type and arm force manipulation state, the user's current training needs and goals are determined, such as strengthening muscle strength and improving joint flexibility. A decision tree algorithm is used to establish a power posture adjustment model, and the fitness action type and training needs are used as input features. The optimal split attribute is selected through indicators such as information gain to generate a decision tree. The output results of the decision tree are used to obtain the optimal power-assisting posture parameters, such as power-assisting direction, power-assisting size, and power-assisting time. The actuators of the exoskeleton or power-assisting device, such as motors and cylinders, are controlled to adjust the power-assisting posture in real time to provide users with personalized fitness assistance. During the training process, the user's arm force manipulation data and fitness movements are continuously monitored at a sampling period of 50 milliseconds, and the newly collected data is added to the training set. The power-assisting posture adjustment model is dynamically updated using an incremental learning algorithm to continuously improve the accuracy and real-time performance of fitness assistance. Through the organic combination of the above steps, intelligent and personalized fitness assistance can be achieved, and the efficiency and effectiveness of fitness training can be improved.

[0011] Step S102, for different types of fitness movements and training requirements, obtain corresponding adjustment strategies from a preset power-assist posture adjustment strategy library, determine the target posture and angle of the power-assist device according to the adjustment strategy, and use the target posture and angle as control parameters for dynamic adjustment of the power-assist device. The power-assist posture adjustment strategy library includes optimal posture parameters for different fitness movements, such as the arm angle during push-ups, the knee angle during squats, etc.

[0012] According to the type of fitness action selected by the user, an adjustment strategy corresponding to the type of fitness action is obtained from a preset power-assisting posture adjustment strategy library; if the user sets a specific training requirement, the obtained adjustment strategy is corrected according to the training requirement to obtain the corrected target posture and angle parameters; the corrected target posture and angle parameters are used as control parameters for dynamic adjustment of the power-assisting device, and the actuator of the power-assisting device is controlled to adjust to the target posture and angle; when the user performs fitness training, the user's motion posture data is collected in real time, and compared with the optimal posture parameters; if there is a deviation between the user's motion posture and the optimal posture, the posture deviation value is calculated through a machine learning algorithm; according to the posture deviation value, the posture and angle of the power-assisting device are dynamically adjusted to correct the user's motion posture in real time; the user's motion posture changes are continuously tracked, and the posture and angle of the power-assisting device are dynamically fine-tuned according to the changes until the user completes the training of the current fitness action.

[0013] Specifically, when the user chooses to perform dumbbell bench press training, the system obtains the adjustment strategy corresponding to the dumbbell bench press from the preset power-assisted posture adjustment strategy library. The strategy includes parameters such as the flexion and extension angle range of the upper limbs and the angle between the trunk and the ground. If the user sets specific training requirements, such as increasing the training intensity, the system calculates a larger flexion and extension angle range of the upper limbs based on the user's body data through the human kinematics model, and adjusts the angle between the trunk and the ground to 30° to form the corrected target posture and angle parameters. The system transmits the corrected parameters to the control unit of the power-assisted device, and the control unit controls the extension and contraction amount and angle of the actuator according to the parameters, so that the power-assisted device is adjusted to the target posture. When the user starts training, the system collects the user's motion posture data in real time through the motion capture device, and compares it with the preset optimal posture parameters. If the user's upper limb flexion and extension angle exceeds the optimal range by more than 5°, the posture deviation value is calculated through the support vector machine algorithm. According to the deviation value, the system dynamically adjusts the extension and contraction amount of the power-assisted device through the PID control algorithm to correct the user's upper limb flexion and extension angle. At the same time, the system continuously tracks the changes in the angle between the user's torso and the ground. When the angle deviates from 30°, the system estimates the angle through the Kalman filter algorithm and controls the power assist device to make corresponding angle adjustments to ensure the correctness of the user's torso posture. After the user completes a set of dumbbell bench press training, the system analyzes the posture data of the entire training process, identifies the key points where the user is prone to posture deviation through the clustering algorithm, and adjusts the power assist strategy in a targeted manner to optimize the auxiliary effect of the next set of training.

[0014] Step S103, obtain the real-time posture and angle information of the power assist device, compare it with the target posture and angle, calculate the deviation value of the posture and angle, and determine whether the posture of the power assist device needs to be adjusted according to the deviation value. The calculation formula of the deviation value is: deviation value = actual value - target value.

[0015] The real-time posture and angle information of the power-assisting device are obtained through the sensor, and the obtained real-time posture and angle information are recorded as the actual posture and the actual angle respectively. The target posture and target angle of the current power-assisting device are obtained from the preset target posture and target angle parameters. According to the deviation value calculation formula, the posture deviation value between the actual posture and the target posture, and the angle deviation value between the actual angle and the target angle are calculated respectively. It is judged whether the calculated posture deviation value exceeds the preset posture deviation threshold. If it exceeds the threshold, it is determined that the posture of the power-assisting device needs to be adjusted; if it does not exceed the threshold, the posture of the power-assisting device does not need to be adjusted. It is judged whether the calculated angle deviation value exceeds the preset angle deviation threshold. If it exceeds the threshold, it is determined that the angle of the power-assisting device needs to be adjusted; if it does not exceed the threshold, the angle of the power-assisting device does not need to be adjusted. If at least one of the posture deviation value and the angle deviation value is judged to need to be adjusted, the PID control algorithm is used to calculate the adjustment amount of the posture and angle of the power-assisting device according to the posture deviation value and the angle deviation value, and the adjustment amount is sent to the actuator of the power-assisting device. The actuator adjusts the posture and angle of the power-assisting device accordingly according to the received adjustment amount, until the actual posture and actual angle are respectively consistent with the target posture and target angle, thus completing the posture adjustment process of the power-assisting device.

[0016] Specifically, the real-time posture and angle information of the power-assisting device are obtained through the six-axis sensor, and the obtained real-time posture and angle information are recorded as the actual posture and actual angle respectively. The preset target posture and target angle parameters are read from the controller memory to obtain the target posture and target angle of the current power-assisting device. According to the deviation value calculation formula, the Euler angle deviation value between the actual posture and the target posture, and the angle deviation value between the actual angle and the target angle are calculated respectively. It is judged whether the calculated Euler angle deviation value exceeds the preset 2° threshold value. If it exceeds the threshold value, it is determined that the posture of the power-assisting device needs to be adjusted; if it does not exceed the threshold value, the posture of the power-assisting device does not need to be adjusted. At the same time, it is judged whether the calculated angle deviation value exceeds the preset 5° threshold value. If it exceeds the threshold value, it is determined that the angle of the power-assisting device needs to be adjusted; if it does not exceed the threshold value, the angle of the power-assisting device does not need to be adjusted. If the posture or angle deviation exceeds the threshold value, the PID control algorithm is used to calculate the adjustment amount of the posture and angle of the power-assisting device according to the deviation value, wherein the proportional coefficient is 5, the integral coefficient is 5, and the differential coefficient is 2, and the adjustment amount is sent to the actuator of the power-assisting device through the CAN bus. The servo motor inside the actuator drives the gear transmission mechanism to drive the joints of the power-assisting device to make corresponding adjustments based on the received adjustment amount. The sensor monitors and feedbacks the adjusted posture and angle in real time, and continuously corrects them through closed-loop control until the deviation values ​​of the actual posture and actual angle from the target posture and target angle respectively are less than the set threshold, completing the posture adjustment process of the power-assisting device and realizing human-machine collaboration.

[0017] Step S104, determining whether the deviation value exceeds a preset threshold value, and if the deviation value exceeds the preset threshold value, adjusting the actuator, wherein the deviation value is the adjustment amount of the actuator.

[0018] In one embodiment, if the deviation value exceeds a preset threshold, the dynamic adjustment mechanism of the power assist device is triggered, and the actuator of the power assist device is controlled to make corresponding posture and angle adjustments according to the positive and negative and size of the deviation value until the actual posture and angle are consistent with the target value. The actuator includes power devices such as motors and hydraulic cylinders, and precise adjustment is achieved through a PID control algorithm.

[0019] The current actual posture and actual angle data of the power-assisting device are obtained, and compared with the preset target posture and target angle respectively, and the posture deviation value and angle deviation value are calculated. It is judged whether the posture deviation value and angle deviation value exceed the preset deviation threshold value. If they exceed the threshold value, the dynamic adjustment mechanism of the power-assisting device is triggered; if they do not exceed the threshold value, the current posture and angle are maintained unchanged. When the dynamic adjustment mechanism is triggered, the direction in which the actuator needs to be adjusted is determined according to the positive and negative attributes of the posture deviation value and the angle deviation value, that is, whether the current posture and angle need to be increased or decreased. According to the size attributes of the posture deviation value and the angle deviation value, the PID control algorithm is used to calculate the specific amplitude that the actuator needs to adjust, so as to achieve precise adjustment of the posture and angle of the power-assisting device. The motor device or hydraulic device in the actuator is controlled to output the corresponding current or hydraulic command according to the calculation result of the PID algorithm, drive the motor or hydraulic cylinder to produce the corresponding action, and adjust the posture and angle of the power-assisting device. After the actuator completes an adjustment action, the actual posture and actual angle data of the power-assisting device are obtained again, and the deviation value between it and the target posture and target angle is calculated to determine whether the deviation is within an acceptable range. If the deviation still exceeds the acceptable range, the step of determining whether the posture deviation value and the angle deviation value exceed the preset deviation threshold is executed according to the updated deviation value, until the actual posture and actual angle of the power assist device are consistent with the target posture and target angle, completing the dynamic adjustment process.

[0020] Specifically, the power assist device obtains the current actual posture and actual angle data in real time through the sensor. For example, the current posture is (10°, 20°, 30°) and the actual angle is (5°, 15°, 25°). It is compared with the preset target posture (0°, 0°, 0°) and target angle (0°, 0°, 0°), and the posture deviation value (10°, 20°, 30°) and angle deviation value (5°, 15°, 25°) are calculated. The system determines that the posture deviation value and the angle deviation value exceed the preset deviation threshold value ±2°, triggering the dynamic adjustment mechanism of the power assist device. According to the fact that the posture deviation value and the angle deviation value are both positive values, it is determined that the actuator needs to adjust the posture and angle in the negative direction. Using the PID control algorithm, the proportional coefficient Kp=5, the integral coefficient Ki=1, and the differential coefficient Kd=2 are set to calculate the specific amplitude that the actuator needs to adjust, and output the control command of the actuator, such as the motor rotates -15° or the hydraulic cylinder extends -20mm, to drive the power assist device to move and adjust its posture and angle. After completing an adjustment, the sensor reacquires the actual posture (2°, 3°, 5°) and actual angle (1°, 2°, 3°) of the power assist device. The calculated deviation value still exceeds the acceptable range of ±1° of the target value. According to the updated deviation value, the above control process is repeated until the actual posture and angle are consistent with the target value, completing the dynamic adjustment. The entire process realizes the posture and angle adjustment of the power assist device through closed-loop control, improving its motion accuracy and stability.

[0021] In one embodiment, the steps are also included: during the dynamic adjustment process of the power-assisting device, the user's arm strength control changes are monitored in real time to determine whether the user's fitness movements have changed. If a movement change is detected, the user's arm strength control data is re-acquired, the movement type and training requirements are analyzed, and the posture adjustment strategy of the power-assisting device is updated accordingly.

[0022] The user's arm force manipulation data is collected in real time through sensors, and the data is transmitted to the control unit for analysis and processing. The control unit extracts features and recognizes patterns on the collected arm force manipulation data according to the preset action model library to determine the user's current fitness action type. If the user's fitness action is identified to have changed, the posture adjustment mechanism of the power-assisting device is triggered, and the posture adjustment strategy of the power-assisting device is determined according to the change of the action type. The control unit converts the posture adjustment strategy into a control instruction, transmits it to the actuator, drives the joint motor of the power-assisting device to make corresponding angle adjustments, and realizes the dynamic optimization of the posture of the power-assisting device. During the posture adjustment process, the user's arm force manipulation changes are continuously monitored through the force feedback sensor, and the feedback data is transmitted to the control unit to form a real-time closed-loop control. The control unit dynamically adjusts the resistance size and auxiliary force of the power-assisting device according to the force feedback data and the preset training demand model to ensure that the user obtains the appropriate training intensity during exercise. If the user's fitness action is detected to have changed again during the continuous monitoring process, the above steps are repeated to trigger a new round of posture adjustment and power optimization, and realize the real-time synchronization of the power-assisting device and the user's action.

[0023] Specifically, the sensor collects the user's arm force manipulation data at a frequency of 50Hz, including torque information in three-dimensional space. The control unit uses the support vector machine (SVM) algorithm to extract features and recognize patterns from the collected data. By calculating the similarity between the data and each action template in the action model library, the action type with the highest similarity is selected as the judgment result. When the similarity exceeds 8, it is determined that the user's fitness action has changed. The control unit selects the corresponding strategy from the preset posture adjustment strategy library according to the change of the action type, and converts the strategy into the angle adjustment value of each joint motor by solving the inverse kinematics equation. The angle accuracy is controlled within 1°. The actuator drives the motor to move according to the control instruction and completes the posture adjustment within 200 milliseconds. The force feedback sensor monitors the user's arm force manipulation changes at a frequency of 100Hz, and filters the data through the Kalman filter algorithm to extract the change trend of the force. The control unit compares the force feedback data with the training demand model, and adjusts the resistance and auxiliary force of the power assist device through the fuzzy control algorithm. The resistance adjustment range is 0-50N, the auxiliary force adjustment range is 0-100N, and the adjustment accuracy is 1N. When the force feedback data deviates from the training demand model by more than 10%, the resistance and auxiliary force adjustment is triggered. Through real-time closed-loop control, it ensures that users get the appropriate training intensity during exercise and improves fitness effects.

[0024] In one embodiment, the steps are also included: by evaluating the posture adjustment effect of the power-assisting device, obtaining objective data such as the joint angle and torque of the power-assisting device, optimizing and updating the reflex lever model and the power-assisting posture adjustment strategy library, and continuously improving the dynamic adaptability of the power-assisting device to better match the mechanical parameters of different users.

[0025] Obtain objective data such as joint angle and torque of the power-assisting device as input for evaluating the posture adjustment effect. According to the objective data obtained, the posture adjustment effect of the power-assisting device is evaluated by using a support vector machine algorithm to obtain an evaluation result. If the evaluation result does not reach the preset threshold, the evaluation result is used as input, and the reflection arm model is optimized by a decision tree algorithm to obtain an optimized reflection arm model. The optimized reflection arm model is used as input, and the power-assisting posture adjustment strategy library is updated by using a reinforcement learning algorithm to obtain an updated power-assisting posture adjustment strategy library. According to the updated power-assisting posture adjustment strategy library, the posture adjustment parameters of the power-assisting device for the next time are determined, and applied to the control of the power-assisting device. Obtain objective data such as joint angle and torque of the power-assisting device after applying the posture adjustment parameters, and determine whether it matches the mechanical parameters of different users. If the matching degree does not reach the preset threshold, the newly obtained objective data is used as input, and the step of optimizing the reflection arm model by a decision tree algorithm to obtain the optimized reflection arm model is performed until the dynamic adaptability of the power-assisting device reaches the expected requirements.

[0026] Specifically, objective data such as joint angle and torque of the power-assist device are obtained through sensors, such as joint angle in the range of 0-90° and torque in the range of 0-50N·m. The support vector machine algorithm is used to evaluate the data, and the data is mapped to a high-dimensional space through the Gaussian kernel function to construct the optimal classification hyperplane, and the posture adjustment effect score is 85. If the score does not reach the preset threshold of 9, it is input into the decision tree algorithm, and the feature with the largest information gain is selected for branching through the ID3 algorithm to optimize the length, mass, damping and other parameters of each joint of the reflex arm model. The optimized model is input into the reinforcement learning algorithm, and the actions are continuously tried through Q-learning and the strategy is updated according to the feedback to obtain new power-assist postures such as elbow flexion 90° and shoulder abduction 30°. The new parameters are applied to the control, and the data is obtained again to determine its matching degree with different users such as 7m in height and 70kg in weight. If it is less than 95, it is iterated and optimized until the dynamic adaptation performance reaches the expected level.

[0027] The present invention also provides a power assist device, including a controller for executing a method as described in any of the above embodiments; an actuator, including a motor 1, a first connecting arm 2 and a second connecting arm 3, and the motor 1 is respectively connected to the controller, the first connecting arm 2 and the second connecting arm 3.

[0028] See also Figure 2 , Figure 2 It is a schematic diagram of the actuator of the present invention, the motor 1 includes a reduction motor 1 and a stepper motor 1, the first connecting arm 2 and the second connecting arm 3 form an angle, the motor 1 is connected to the first connecting arm 2 and the second connecting arm 3 respectively, the angle between the first connecting arm 2 and the second connecting arm 3 can be increased or decreased by the motor 1, in one embodiment, the number of actuators is multiple, and the second connecting arm 3 of the previous actuator serves as the first connecting arm 2 of the next actuator. In this embodiment, after the user wears the power-assisting device, the motor 1 is installed at the joint through the first connecting member 11, the first connecting arm 2 and the second connecting arm 3 are installed at the upper arm and the lower arm respectively through the second connecting member 21 and the third connecting member 31, the first connecting member 11, the second connecting member 21 and the third connecting member 31 include protective gear for assembly on the user. In one embodiment, a gyroscope sensor and an acceleration sensor are also included, and the user's arm strength control data is obtained in real time through the sensors, and the arm strength control data is compared with the target posture and angle. For example, during weightlifting exercise, the angle between the upper arm and the forearm is the angle between the first connecting arm 2 and the second connecting arm 3. If the angle between the upper arm and the forearm is not less than the target angle, it is determined to be a non-standard posture. The motor 1 drives the first connecting arm 2 and the second connecting arm 3 to assist the user in contracting the angle between the upper arm and the forearm until the angle between the upper arm and the forearm is less than the target angle, thereby preventing the user's arm from being tired and losing strength, and matching the user's exercise movements with the target fitness movements and training needs.

[0029] In another embodiment, the power assist device can be switched between power assist and resistance modes. For example, if user A has difficulty completing a standard pull-up, the power assist mode of the power assist device can be activated, and the power assist device assists the user to shrink the angle between the upper arm and the forearm, so that user A can complete the pull-up action. For another example, if user B wishes to increase the intensity of arm strength during weightlifting, the resistance mode of the power assist device can be activated, and the power assist device hinders the user from shrinking the angle between the upper arm and the forearm, so that user B can increase the intensity of arm strength. On the one hand, the power assist device can generate a force in the same direction as the user's exercise, so as to assist the user in exercising and avoid fatigue and weakness. On the other hand, it can generate a force in the opposite direction of the user's exercise, so as to increase the intensity of the user's exercise. It should be noted that the specific principle of switching the working mode of the power assist device through the switch can be determined by referring to existing materials, and the present invention does not make specific limitations on this.

[0030] The above embodiments are only used to illustrate the technical solution of the present invention but not to limit it. The present invention is described in detail with reference to the preferred embodiments. It should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention, and should be included in the scope of the claims of the present invention.

Claims

1. A fitness assistance method, characterized in that: The power assist device includes a controller and an actuator. The method is applied to the controller. The method includes: Acquire the user's arm force manipulation data, and analyze the arm force manipulation data based on a pre-established reflex arm model to obtain the user's fitness action type and training needs; wherein the arm force manipulation data includes the joint angle and torque of the actuator; According to the fitness action type and training requirements, a corresponding adjustment strategy is obtained from a preset power-assisting posture adjustment strategy library to determine a target posture and angle of the power-assisting device; Acquire the real-time posture and angle information of the power assist device, compare it with the target posture and angle, and calculate the deviation value of the posture and angle; Determine whether the deviation value exceeds a preset threshold value, and if the deviation value exceeds the preset threshold value, adjust the actuator, wherein the deviation value is the adjustment amount of the actuator.

2. The fitness assistance method according to claim 1, characterized in that: The acquiring of the user's arm strength manipulation data and analyzing the arm strength manipulation data based on a pre-established reflex arm model to obtain the user's fitness action type and training needs include: The user's lever arm length, mass distribution, and joint friction coefficient are collected as parameters by sensors, and the collected parameters are input into the reflex lever arm model; The user's arm force control data is calculated using a dynamic equation to obtain the user's current arm force control state; Determine the user's current fitness action type by performing feature extraction and pattern recognition on the arm force manipulation data; Determine the user's current training needs based on the user's fitness action type and arm strength control status.

3. The fitness assistance method according to claim 1, characterized in that: The step of obtaining a corresponding adjustment strategy from a preset assisting posture adjustment strategy library according to the fitness action type and training requirements includes: According to the fitness action type selected by the user, a corresponding adjustment strategy is obtained from the assisting posture adjustment strategy library; If the user sets a specific training requirement, the acquired adjustment strategy is corrected according to the training requirement to obtain the corrected target posture and angle parameters.

4. The fitness assistance method according to claim 1, characterized in that: The step of acquiring the real-time posture and angle information of the power assisting device, comparing the real-time posture and angle information with the target posture and angle, and calculating the deviation value of the posture and angle includes: Acquire real-time posture and angle information of the power assist device through sensors; The acquired real-time posture and angle information are compared with the target posture and target angle respectively, and the posture deviation value and the angle deviation value are calculated.

5. The fitness assistance method according to claim 1, characterized in that: The determining whether the deviation value exceeds a preset threshold value, and if the deviation value exceeds the preset threshold value, adjusting the actuator, comprises: Determine whether the posture deviation value and the angle deviation value exceed a preset deviation threshold; If it exceeds, the PID control algorithm is used to calculate the adjustment amount of the posture and angle of the power assist device according to the positive and negative and size of the posture deviation value and the angle deviation value, and the actuator is controlled to make adjustments.

6. The fitness assistance method according to claim 5, characterized in that: The control actuator is adjusted, including: The user's arm strength control changes are monitored in real time. If a change in fitness movement is detected, the user's arm strength control data is re-acquired and the posture adjustment strategy of the power assist device is updated.

7. The fitness assistance method according to claim 6, characterized in that: The real-time monitoring of the user's arm force manipulation changes includes: Collect the user's arm force manipulation data in real time through sensors; According to the preset action model library, feature extraction and pattern recognition are performed on the collected arm force manipulation data to determine the user's current fitness action type; If a change in the user's fitness movements is detected, an alarm is issued.

8. The fitness assistance method according to claim 1, characterized in that: The method further comprises: The posture adjustment effect of the power-assisting device is evaluated, and the reflection lever arm model and the power-assisting posture adjustment strategy library are optimized and updated according to the evaluation results.

9. The fitness assistance method according to claim 8, characterized in that: The evaluating of the posture adjustment effect of the power assisting device and optimizing and updating the reflection lever arm model and the power assisting posture adjustment strategy library according to the evaluation result include: Obtain objective data such as joint angle and torque of the power assist device, and use support vector machine algorithm to evaluate the posture adjustment effect; If the evaluation result does not reach the preset threshold, the decision tree algorithm is used to optimize the reflection arm model, and the reinforcement learning algorithm is used to update the power-assisted posture adjustment strategy library.

10. A power assist device, characterized in that: include: A controller, configured to execute the method according to any one of claims 1 to 9; The actuator comprises a motor, a first connecting arm and a second connecting arm, wherein the motor is connected to the controller, the first connecting arm and the second connecting arm respectively.

Images