A gluteal muscle training self-adaptive torque compensation method based on pelvic compensation feedback

Abstract

The application belongs to the technical field of fitness equipment, and particularly relates to a hip muscle training self-adaptive torque compensation method based on pelvic compensation feedback, which comprises the following steps: acquiring real-time pedaling displacement in a centripetal abduction stage in a unilateral hip muscle training process, combining with a pre-stored maximum pedaling length of a user, calculating a real-time displacement ratio, and determining a force arm compensation coefficient; acquiring a reference average speed of a first action in the centripetal abduction stage and a current average speed of a current action, calculating a speed retention rate, and adjusting a speed fatigue adjustment coefficient; acquiring an initial inclination angle of a pelvis when the user is in a body upright neutral position, and a current pelvic inclination angle and a current inclination angle in a training process; calculating a posture penalty coefficient; and dynamically adjusting a preset initial resistance by using the speed fatigue adjustment coefficient, the posture penalty coefficient and the force arm compensation coefficient. The application scheme can realize real-time sensing of human kinematics data and spatial posture, and accordingly realize dynamic torque compensation and multi-modal safety intervention.

Description

Technical Field

[0001] This application relates to the field of fitness equipment technology, specifically to an adaptive torque compensation method for gluteal muscle training based on pelvic compensatory feedback. Background Technology

[0002] The gluteus maximus is a crucial component for core stability and lower limb power generation. Strengthening the gluteal muscles plays a central role in sports rehabilitation, posture correction, and general fitness. Currently, common glute training equipment on the market mainly consists of traditional mechanical weighted multi-gyms (such as cable crossover machines and low back pull-up machines) or resistance band equipment. However, with the development of sports biomechanics and rehabilitation medicine, traditional training equipment suffers from the following technical shortcomings: The physical resistance pattern is singular and severely disconnected from the body's physiological torque. Traditional weighted equipment relies on gravity to provide constant resistance, while the resistance of elastic bands increases linearly with the stretching distance. However, when the gluteus maximus performs concentric abduction or extension movements, its physiological torque curve exhibits a non-linear characteristic: in the later stages of the movement (peak contraction zone), the muscle is in its shortest contraction position, prone to "insufficient active force," at which point the muscle torque drops sharply, easily leading to cramps or spasms. The traditional constant or linearly increasing resistance pattern cannot dynamically and smoothly compensate for changes in physiological torque, resulting in insufficient stimulation at the beginning of the movement and difficulty in completing the standard movement at the end due to excessive resistance.

[0003] The lack of posture monitoring and anti-compensation mechanisms can easily lead to lumbar spine injuries. When users face excessive loads or enter a period of fatigue, they often distort their movements to complete the set range of motion. Typical incorrect compensatory behaviors include using methods such as anterior pelvic tilt and lower back collapse to gain leverage. Traditional equipment cannot sense changes in the user's spatial posture and continues to output high damping forces even when severe lumbar compensation occurs. This causes shear forces to act directly on the lumbar intervertebral discs, weakening the training effect and increasing the risk of sports injuries.

[0004] The lack of intelligent fatigue reduction logic can easily lead to movement breakdown. In a set of continuous training, muscle explosive power and contraction speed decrease non-linearly with the depletion of physical energy. Traditional weight machines maintain a constant resistance throughout the set. When trainees experience deep fatigue, they often have to resort to sacrificing proper form or stopping the exercise altogether, failing to achieve scientific, safe, and high-quality exhaustion stimulation.

[0005] The inability to correct bilateral asymmetry exacerbates muscle imbalances. Many individuals experience left-right limb strength imbalances. When using traditional equipment for unilateral alternating training, the lack of precise quantitative tracking of muscle fatigue slope and work quality often leads users to unconsciously exert more control on the stronger leg while neglecting the weaker one. Traditional equipment cannot distinguish between the supporting and exerting legs, nor can it provide differentiated resistance prescriptions at the physical level, resulting in a "stronger leg remaining strong, weaker leg remaining weak" situation, making it difficult to achieve rehabilitative correction of bilateral muscle strength.

[0006] In summary, existing glute training devices suffer from technical deficiencies such as rigid resistance modes, lack of closed-loop posture monitoring, inability to adapt to fatigue states, and lack of mechanisms to correct muscle asymmetry. Therefore, there is an urgent need in this field for an intelligent training control method capable of real-time sensing of human kinematic data and spatial posture, and accordingly performing dynamic torque compensation and multimodal safety intervention. Summary of the Invention

[0007] In order to perceive human kinematic data and spatial posture in real time, and to perform dynamic torque compensation and multimodal safety intervention accordingly, in a first aspect, embodiments of this application provide an adaptive torque compensation method for gluteal muscle training based on pelvic compensatory feedback, the method comprising: The real-time push-off displacement during the concentric abduction phase of unilateral glute training is obtained, and the real-time displacement ratio is calculated by combining it with the pre-stored maximum push-off length of the user. The lever arm compensation coefficient is determined based on the real-time displacement ratio. Obtain the baseline average velocity of the first movement and the current average velocity of the current movement during the concentric abduction phase, and calculate the velocity retention rate; The speed fatigue adjustment coefficient is adjusted stepwise according to the speed retention rate. The system obtains the initial pelvic pitch angle when the user is in an upright and neutral position, as well as the current pelvic tilt angle and current pitch angle during the training process, and determines the pitch angle change based on the current pitch angle and the initial pitch angle. If the current pelvic tilt angle does not exceed the preset safe tilt range and the pitch angle change does not exceed the preset safe pitch change threshold, the attitude penalty coefficient is set to 0; otherwise, the attitude penalty coefficient is calculated using the following two-dimensional proportional feedback model. : ; in, For roll penalty gain, For pitch penalty gain, As the maximum penalty limit, This represents the current pelvic tilt angle. This represents the change in pitch angle. To pre-set a safe roll range, To preset a safe pitch change threshold; Using speed fatigue adjustment coefficient Attitude penalty coefficient And lever arm compensation coefficient The initial resistance is preset based on the current training mode. F Dynamic adjustments are made to obtain the target damping force for the next action. T To perform dynamic resistance control, the updated model is as follows: .

[0008] In one possible implementation, the maximum push-off length is obtained through the following steps: Monitor a single concentric stretching motion of the lower limbs under baseline damping force; When the change in pelvic pitch angle exceeds the preset safe pitch change threshold, the push-off displacement at this time is obtained as the maximum push-off length. The step of determining the lever arm compensation coefficient based on the real-time displacement ratio includes: when At that time, the lever arm compensation coefficient ; when At that time, the lever arm compensation coefficient ; when At that time, the lever arm compensation coefficient ; in The preset peak safety factor, This is the upper limit of the initial power zone. This represents the lower limit of the peak contraction zone. This represents the real-time displacement ratio.

[0009] In one possible implementation, the stepwise adjustment of the speed fatigue adjustment coefficient based on the speed retention rate includes: when At that time, speed fatigue adjustment coefficient ; when At that time, speed fatigue adjustment coefficient ; when At that time, speed fatigue adjustment coefficient ; in, R For speed retention rate, R 1 represents the first fatigue threshold. R 2 represents the second fatigue threshold. For the attenuation slope, This represents the minimum safety factor.

[0010] In one possible implementation, the method further includes: During continuous training of the same gluteal muscles, record the number of repetitions N and determine the number of repetitions. The average speed of movement during the concentric abduction phase of this movement Calculate the average value of the action sequence. and the overall average speed The fatigue descent slope of the current training side was calculated using the least squares method. : ; ; ; The gluteal muscles on the currently trained side are determined based on the vector direction of the current pelvic tilt angle, combined with the fatigue descent slope. K Identify the weaker gluteal muscle and automatically increase the target damping force during the eccentric phase for the weaker gluteal muscle in subsequent training.

[0011] In one possible implementation, the automatic addition of the target damping force during the eccentric phase to the weaker gluteal muscle in subsequent training includes: During the centripetal extension phase where v > 0, the target damping force is: T During the centrifugal recovery phase when v < 0, the target damping force is... T Multiply by centrifugal overload factor O ecc As the actual output damping force, among which O ecc >1.

[0012] In one possible implementation, the method further includes: The damping force output by the equipment during the period of peak contraction [ t 1, t Integrating within 2], we obtain the peak tension impulse for a single instance used to assess the depth of gluteal muscle stimulation. I : ; Among them, the push-off displacement is not less than c× M The dynamic range is the peak contraction region, and c is a preset constant proportional coefficient; Output the glute mass control index after training. G Its calculation model is as follows: ; ; in, N The number of times the action is completed. For the first The effective range of motion for each movement. The target tension impulse preset by the system, For the first The posture penalty coefficient triggered by the next action. , , All of these are preset weighting coefficients.

[0013] Secondly, this application provides an adaptive torque compensation gluteal muscle training device, wherein the training device performs any of the above-mentioned adaptive torque compensation methods for gluteal muscle training based on pelvic compensation feedback during use. The training device includes: a housing, an attitude sensing acquisition device, a flexible pad, a sliding plate, a rack, a slide rail, a rotating shaft, and gears; The rotating shaft is rotatably connected to the housing and is connected to the gear. One end of the rack is located inside the housing and meshes with the gear. The other end of the rack is located outside the housing and is connected to the slide plate. The slide plate is slidably connected to the slide rail. The flexible pad is disposed on the slide plate. The attitude sensing acquisition device is adapted to be worn on the user's pelvic area and is configured to collect pelvic attitude data in real time. The pelvic attitude data includes tilt angle, pitch angle, and lateral acceleration.

[0014] In one possible implementation, the training device further includes: a skateboard base, a first bracket, a magnetic powder brake, a second bracket, a rotary encoder, a front rack retainer, a slide, a slider, a crank, and a rear rack retainer. Both the first bracket and the second bracket are disposed within the housing. The magnetic powder brake is disposed on the first bracket, and the rotary encoder is disposed on the second bracket. The rotating shaft is rotatably connected between the first bracket and the second bracket. One end of the rotating shaft is connected to the magnetic powder brake, and the other end is connected to the rotary encoder. The rotary encoder is configured to detect the rotation of the gear in real time to calculate the movement speed and pedal displacement. The portion of the rack located inside the housing is supported by the front rack retainer and the rear rack retainer; the end of the rack located outside the housing is provided with a crank and a slider, the slider is disposed in a groove on the slide rail, and the slider is connected to the slide plate and the slide plate base.

[0015] In one possible implementation, the training device further includes a touch control module for displaying preset gluteal muscle training modes.

[0016] Thirdly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements any of the above-mentioned adaptive torque compensation methods for gluteal muscle training based on pelvic compensation feedback.

[0017] This application provides an adaptive torque compensation method for gluteal muscle training based on pelvic compensation feedback, comprising: acquiring the real-time push-off displacement during the concentric abduction phase of unilateral gluteal muscle training, and calculating the real-time displacement ratio by combining it with the pre-stored maximum push-off length of the user; determining the lever arm compensation coefficient based on the real-time displacement ratio; acquiring the baseline average speed of the first movement and the current average speed of the current movement during the concentric abduction phase, and calculating the speed maintenance rate; adjusting the speed fatigue adjustment coefficient in a stepwise manner based on the speed maintenance rate; acquiring the initial pelvic pitch angle when the user is in an upright and neutral position, and the current pelvic tilt angle and current pitch angle during the training process, and determining the pitch angle change based on the current pitch angle and the initial pitch angle; setting the posture penalty coefficient to 0 if the current pelvic tilt angle does not exceed a preset safe tilt range and the pitch angle change does not exceed a preset safe pitch change threshold; otherwise, calculating the posture penalty coefficient using the following two-dimensional proportional feedback model. : ; in, For roll penalty gain, For pitch penalty gain, As the maximum penalty limit, This represents the current pelvic tilt angle. This represents the change in pitch angle. To pre-set a safe roll range, To preset a safe pitch change threshold; Using speed fatigue adjustment coefficient Attitude penalty coefficient And lever arm compensation coefficient The initial resistance is preset based on the current training mode. F Dynamic adjustments are made to obtain the target damping force for the next action. T To perform dynamic resistance control, the updated model is as follows: .

[0018] The proposed solution can sense human kinematic data and spatial posture in real time, and perform dynamic torque compensation and multimodal safety intervention accordingly. Attached Figure Description

[0019] Figure 1 A schematic diagram of a method for adaptive torque compensation in gluteal muscle training based on pelvic compensation feedback is provided for an embodiment of this application. Figure 2 A schematic diagram of a first structure of an adaptive torque compensation gluteal muscle training device provided in an embodiment of this application; Figure 3 A second structural schematic diagram of an adaptive torque compensation gluteal muscle training device provided in an embodiment of this application; Figure 4 This is a schematic diagram of a third structure of an adaptive torque compensation gluteal muscle training device provided in an embodiment of this application.

[0020] In the diagram, 1 is the housing, 2 is the touch control module, 3 is the attitude sensor, 4 is the flexible pad, 5 is the skateboard, 6 is the skateboard base, 7 is the first bracket, 8 is the magnetic powder brake, 9 is the left rear swivel wheel, 10 is the left front swivel wheel, 11 is the right front swivel wheel, 12 is the right rear swivel wheel, 13 is the base plate, 14 is the second base plate, 15 is the second bracket, 16 is the rotary encoder, 17 is the front rack retainer, 18 is the rack, 19 is the front rack bracket, 20 is the rear rack bracket, 21 is the slide groove, 22 is the slider, 23 is the sliding knee, 24 is the crank, 25 is the rear rack retainer, 26 is the rotating shaft, and 27 is the gear. Detailed Implementation

[0021] In a first aspect, embodiments of this application provide an adaptive torque compensation method for gluteal muscle training based on pelvic compensatory feedback, the method comprising: S101 acquires the real-time push-off displacement during the concentric abduction phase of unilateral gluteal muscle training, and calculates the real-time displacement ratio by combining it with the pre-stored maximum push-off length of the user.

[0022] In this embodiment, only one side of the gluteal muscles is trained each time. The concentric abduction phase refers to the user pushing the leg back, and there is also an eccentric recovery phase, which is the process of the user pulling the leg back.

[0023] The maximum push-off length is obtained through the following steps: Monitor a single concentric stretching motion of the lower limbs under baseline damping force; When the change in pelvic pitch angle exceeds the preset safe pitch change threshold, the push-off displacement at this time is obtained as the maximum push-off length.

[0024] During the personalized calibration phase before formal training, the system outputs a basic resistance of 50N to guide the user in performing a single maximum stroke of centripetal stretching.

[0025] During this process, the system monitors the real-time changes in the pelvic pitch angle. In this embodiment, a preset safe pitch change threshold is used. When the system detects Reaching 5.1° (i.e.) When the system determines that the user's hip extension has reached its current physiological limit of flexibility, it begins to compensate with anterior pelvic tilt. At this point, the system immediately truncates the current rack displacement (e.g., x =0.8m), and recorded it as the user's exclusive maximum push-out length. M =0.8m.

[0026] The real-time push-off displacement is obtained through the following steps: The pulse signals generated by gear rotation are captured by a rotary encoder installed inside the training device; Cumulative pulse count n Based on the resolution of the rotary encoder P and the pitch circle radius of the gear r Calculate the real-time linear displacement of the rack. The rack displacement is the real-time pedal displacement.

[0027] Furthermore, the sampling period can be set. Obtain the displacement change within two adjacent sampling periods. Instantaneous velocity is calculated using a differential algorithm. : ; The average velocity used subsequently is obtained by integrating this instantaneous velocity.

[0028] Real-time displacement ratio p = x / 0.8.

[0029] S102, determine the lever arm compensation coefficient based on the real-time displacement ratio.

[0030] when At that time, the lever arm compensation coefficient ; when At that time, the lever arm compensation coefficient ; when At that time, the lever arm compensation coefficient ; in The preset peak safety factor, This is the upper limit of the initial power zone. This represents the lower limit of the peak contraction zone. This represents the real-time displacement ratio.

[0031] Based on human anatomy, the upper limit of the initial force exertion zone is preset. =0.45 (i.e., 45% of the stroke), lower limit of the peak contraction zone. =0.80 (i.e., 80% of the journey), and set the peak safety factor. =0.60.

[0032] The system establishes the following electronic cam curve: (1) When 0≤ p When ≤0.45, set u ( x =1.0, apply a full-load resistance of 200N; (2) When 0.45 < p When the coefficient is <0.80, the gluteus maximus becomes the dominant force-generating zone, and the resistance coefficient decreases smoothly. The algorithm is as follows: ; (3) When At this point, the gluteus maximus enters the peak contraction zone, at which time it is highly susceptible to spasm, and the system locks the resistance coefficient at [value missing]. =0.6. At this point, the target resistance drops to 200N × 0.6 = 120N.

[0033] S103, obtain the baseline average speed of the first movement and the current average speed of the current movement in the concentric abduction phase, and calculate the speed retention rate.

[0034] Assuming the user's first action reference speed The speed measured for the tenth action is currently... The speed retention rate is calculated as 0.7 / 1.0 = 0.7.

[0035] S104, adjust the speed fatigue adjustment coefficient in a stepwise manner according to the speed retention rate.

[0036] when At that time, speed fatigue adjustment coefficient ; when At that time, speed fatigue adjustment coefficient ; when At that time, speed fatigue adjustment coefficient ; in, R For speed retention rate, R 1 represents the first fatigue threshold. R 2 represents the second fatigue threshold. For the attenuation slope, This represents the minimum safety factor.

[0037] when When the fatigue threshold is not triggered, it is determined that the fatigue threshold has not been triggered. At that time, it was determined to be moderate fatigue; when At that time, it was determined to be severe fatigue.

[0038] In this embodiment, a first fatigue threshold is preset. R 1 = 0.9, the second fatigue threshold R 2 = 0.6, attenuation slope =1.5, minimum safety factor =0.5. Wherein These are the "control parameters (adjustment gain)" pre-programmed into the system's control program. They guide the system on "how to unload stress." When moderate fatigue occurs, the system utilizes... The decrease in speed is mapped to a decrease in resistance. For example, if the speed drops slightly, the system quickly reduces some of the resistance based on the 1.5 slope, allowing the user to continue the action.

[0039] S105, obtain the initial pelvic pitch angle when the user is in an upright and neutral position, as well as the current pelvic tilt angle and current pitch angle during the training process, and determine the pitch angle change based on the current pitch angle and the initial pitch angle.

[0040] S106, if the current pelvic tilt angle does not exceed the preset safe tilt range and the pitch angle change does not exceed the preset safe pitch change threshold, the attitude penalty coefficient is set to 0; otherwise, the attitude penalty coefficient is calculated using the following two-dimensional proportional feedback model. : ; in, For roll penalty gain, For pitch penalty gain, As the maximum penalty limit, This represents the current pelvic tilt angle. This represents the change in pitch angle. To pre-set a safe roll range, To preset a safe pitch change threshold.

[0041] When the speed of movement is greater than 0 and the displacement is pushed out x Continuously increase, while satisfying the absolute value of the roll angle. And the real-time change in pitch angle In cases where the pelvis is stable and the gluteal muscles are effectively engaged, the posture penalty coefficient is adjusted accordingly. Set to 0; exist or In cases where pelvic tilt is detected as an auxiliary force or lumbar compensatory force due to anterior pelvic tilt, a preset tilt penalty gain is introduced. Pitch penalty gain and maximum penalty limit Calculate the attitude penalty coefficient This instantly reduces the target damping force of the current rack to protect the lumbar spine.

[0042] In this embodiment, the preset safe roll range is [-5°, +5°] (i.e.) S roll =5°), tilt penalty gain =0.1, pitch penalty gain =0.15, maximum penalty limit =0.8.

[0043] During the stretching process (v > 0), if the user experiences severe back collapse due to exhaustion, the real-time tilt angle is measured. =2°, while pitch change =8°.

[0044] at this time Not crossed the boundary, but An out-of-bounds error has occurred. The system is then fitted into a two-dimensional proportional feedback model: ; Calculation .

[0045] At the same time, in determining When this happens, the system generates a posture correction signal, triggering a multimodal alarm to forcibly intervene in the user's incorrect actions. This includes: high-frequency vibration warnings output through a belt worn around the user's waist, visual color warnings output through the touch control module, and voice action correction commands output through the associated speaker.

[0046] S107, utilizing speed fatigue adjustment coefficient Attitude penalty coefficient And lever arm compensation coefficient The initial resistance is preset based on the current training mode. F Dynamic adjustments are made to obtain the target damping force for the next action. T To perform dynamic resistance control, the updated model is as follows: .

[0047] In summary, in p At the instant when the pitch is 0.5° and 8° pitch compensation occurs with moderate fatigue, the final target damping force output by the system is: .

[0048] The calculation result is then expressed by the formula. It is converted into an electric current-driven magnetic powder brake to complete the underlying physical execution.

[0049] in, r The pitch circle radius of the gears inside the training device. kThis is the electromagnetic torque proportionality coefficient of the brake. I 0 represents the preset static friction compensation current for the system.

[0050] This application provides an adaptive torque compensation method for gluteal muscle training based on pelvic compensation feedback, comprising: acquiring the real-time push-off displacement during the concentric abduction phase of unilateral gluteal muscle training, and calculating the real-time displacement ratio by combining it with the pre-stored maximum push-off length of the user; determining the lever arm compensation coefficient based on the real-time displacement ratio; acquiring the baseline average speed of the first movement and the current average speed of the current movement during the concentric abduction phase, and calculating the speed maintenance rate; adjusting the speed fatigue adjustment coefficient in a stepwise manner based on the speed maintenance rate; acquiring the initial pelvic pitch angle when the user is in an upright and neutral position, and the current pelvic tilt angle and current pitch angle during the training process, and determining the pitch angle change based on the current pitch angle and the initial pitch angle; setting the posture penalty coefficient to 0 if the current pelvic tilt angle does not exceed a preset safe tilt range and the pitch angle change does not exceed a preset safe pitch change threshold; otherwise, calculating the posture penalty coefficient using the following two-dimensional proportional feedback model. : ; in, For roll penalty gain, For pitch penalty gain, As the maximum penalty limit, This represents the current pelvic tilt angle. This represents the change in pitch angle. To pre-set a safe roll range, To preset a safe pitch change threshold; Using speed fatigue adjustment coefficient Attitude penalty coefficient And lever arm compensation coefficient The initial resistance is preset based on the current training mode. F Dynamic adjustments are made to obtain the target damping force for the next action. T To perform dynamic resistance control, the updated model is as follows: .

[0051] The proposed solution can sense human kinematic data and spatial posture in real time, and perform dynamic torque compensation and multimodal safety intervention accordingly.

[0052] In one example, during continuous training of the same gluteal muscles, the number of repetitions N is recorded and the number of repetitions is determined. The average speed of movement during the concentric abduction phase of this movement Calculate the average value of the action sequence. and the overall average speed The fatigue descent slope of the current training side was calculated using the least squares method. : ; ; ; The gluteal muscles on the currently trained side are determined based on the vector direction of the current pelvic tilt angle, combined with the fatigue descent slope. K Identify the weaker gluteal muscle and automatically increase the target damping force during the eccentric phase for the weaker gluteal muscle in subsequent training.

[0053] in K It is the "physiological characteristic value (evaluation result)" of a user calculated by the system after monitoring multiple sets of user actions in real time.

[0054] Lateral acceleration or roll angle When the vector direction points to the left, it is determined that the body's center of gravity has shifted to the left, the left leg is the supporting leg, the current training side is identified as the right gluteal muscle, and subsequent training data is included in the right-side database to calculate the fatigue descent slope of the right side. K 2; Lateral acceleration or roll angle When the vector direction points to the right, it is determined that the body's center of gravity has shifted to the right, the right leg is the supporting leg, the current training side is identified as the left gluteal muscles, and subsequent training data is included in the left-side database to calculate the fatigue descent slope of the left side. K 1; Lateral acceleration refers to the acceleration of the human pelvis in the horizontal plane, perpendicular to the front of the body, in the left-right direction. It is used to determine the direction of weight transfer, thereby identifying whether the left or right gluteal muscles are being trained. This is collected in real time by a posture sensor built into a waist belt worn around the pelvis. This sensor is typically a 6-axis or 9-axis MEMS (Micro-Electro-Mechanical System) inertial measurement unit, which can directly output acceleration values ​​in three axes, including lateral acceleration.

[0055] Specifically, during the centripetal extension phase when v > 0, the target damping force is: T During the centrifugal recovery phase when v < 0, the target damping force is... T Multiply by centrifugal overload factor O ecc As the actual output damping force, among which O ecc >1.

[0056] After a series of alternating single-leg training exercises, the system obtained the following result using least squares fitting: the slope of fatigue descent on the left side. K 1 = -0.05, fatigue descent slope on the right side K 2 = -0.02.

[0057] The left leg clearly fatigued faster, leading the system to identify the left side as the weaker gluteal muscle. (Preset eccentric overload coefficient) O ecc =1.2.

[0058] In the next training session, when it is recognized that the left leg is being trained and is in the concentric phase ( When the target damping force is output, the normal target damping force is achieved. T Once the centrifugal recovery phase begins ( The system automatically increases the damping force to T ×1.2, using 120% eccentric overload weight, focus on tearing the left gluteal muscle fibers, forcibly aligning the strength levels of both sides.

[0059] In one example, the bilateral fatigue symmetry index can also be calculated using the following formula. : ; in, K 2 represents the fatigue descent slope on the right side; K 1 represents the fatigue descent slope on the left side.

[0060] In one example, the method further includes: The damping force output by the equipment during the period of peak contraction [ t 1, t Integrating within 2], we obtain the peak tension impulse for a single instance used to assess the depth of gluteal muscle stimulation. I : ; Among them, the push-off displacement is not less than c× M The dynamic range is the peak contraction region, and c is a preset constant proportional coefficient of 0.8; Output the glute mass control index after training. G Its calculation model is as follows: ; ; in, N The number of times the action is completed. For the first The effective range of motion for each movement. The target tension impulse preset by the system, For the first The posture penalty coefficient triggered by the next action. , , All are preset weighting coefficients. Specifically, there are preset... = 120 Ns, preset = 0.4, = 0.6, = 0.5.

[0061] Secondly, embodiments of this application provide an adaptive torque-compensated gluteal muscle training device, such as... Figure 2 and combined Figure 3 , Figure 4 As shown, the training device executes any of the above-mentioned adaptive torque compensation methods for gluteal muscle training based on pelvic compensation feedback during use. The training device includes a housing 1, an attitude sensor acquisition unit 3, a flexible pad 4, a slide plate 5, a rack 18, a slide rail 23, a rotating shaft 26, and a gear 27. The rotating shaft 26 is rotatably connected inside the housing 1 and is connected to the gear 27. One end of the rack 18 is located inside the housing 1 and meshes with the gear 27, while the other end of the rack 18 is located outside the housing 1 and is connected to the slide plate 5. The slide plate 5 is slidably connected to the slide rail 23, and the flexible pad 4 is placed on the slide plate 5. The attitude sensor acquisition unit 3 is adapted to be worn on the user's pelvic area and is configured to collect pelvic attitude data in real time, including tilt angle, pitch angle, and lateral acceleration.

[0062] In this embodiment, the housing 1 serves as the load-bearing component of the training device, used to house the internal transmission structure (gears, shafts, etc.), and can play a protective, fixing, and supporting role.

[0063] The posture sensor acquisition unit 3, as the core component for posture data acquisition, is fitted to the user's pelvic area and features an adjustable strap design (e.g., strap width 20mm, length adjustable between 700-900mm). The inner side of the strap has a skin-friendly cotton pad (2mm thick) to improve wearing comfort and prevent skin pressure during prolonged wear. The posture sensor acquisition unit 3 incorporates a three-axis gyroscope and a three-axis accelerometer, enabling real-time and accurate acquisition of the user's pelvic tilt angle, pitch angle, and lateral acceleration.

[0064] The flexible pad 4 is placed on the skateboard 5 to support the user's legs (during training, the user sits or stands, places their legs on the flexible pad, and performs glute training by pedaling the skateboard). It can be made of skin-friendly, highly elastic material (such as memory foam or silicone), with a thickness of 15-20mm and a non-slip textured surface (texture depth 1mm). This not only cushions the impact of the legs when pedaling, improving training comfort, but also prevents the legs from slipping during training, ensuring training safety.

[0065] This embodiment combines real-time pelvic posture data collected by posture sensors to dynamically adjust training torque, adapting to the user's force exertion and posture changes at different training stages. This avoids excessive torque leading to gluteal muscle strain or insufficient torque leading to ineffective training. At the same time, it corrects the user's pelvic posture, ensuring standard training movements and improving the targeting and safety of training.

[0066] Meanwhile, the attitude sensor is fitted to the pelvic area and can collect core attitude data such as tilt angle, pitch angle, and lateral acceleration in real time and with high sampling accuracy and fast response speed, so as to obtain the user's training posture in a timely manner and provide reliable data support for the accurate adjustment of torque compensation, avoiding torque adjustment deviations caused by inaccurate attitude data collection.

[0067] In some embodiments, such as Figure 2 and combined Figure 3 , Figure 4 As shown, the training device also includes a skateboard base 6, a first support 7, a magnetic powder brake 8, a second support 15, a rotary encoder 16, a front rack retainer 17, a slide groove 21, a slider 22, a crank 24, and a rear rack retainer 25. The first support 7 and the second support 15 are both housed within the housing 1. The magnetic powder brake 8 is mounted on the first support 7, and the rotary encoder 16 is mounted on the second support 15. A rotating shaft 26 is rotatably connected between the first support 7 and the second support 15. One end of the rotating shaft 26 is connected to the magnetic powder brake 8, and the other end is connected to the rotary encoder 16. The rotary encoder 16 is configured to detect the rotation of the gear 27 in real time to calculate the movement speed and pedal displacement. The portion of the rack 18 located within the housing 1 is supported by the front rack retainer 17 and the rear rack retainer 25. The end of the rack 18 located outside the housing 1 is provided with a crank 24 and a slider 22. The slider 22 is located within the slide groove 21 on the slide rail 23, and the slider 22 is connected to the skateboard 5 and the skateboard base 6.

[0068] refer to Figure 3 As shown, the bottom of the box 1 is connected to a base plate 13. The bottom surface of the base plate 13 is provided with a left rear universal wheel 9, a left front universal wheel 10, a right front universal wheel 11, and a right rear universal wheel 12. A second base plate 14 is fixed on the top of the base plate 13.

[0069] In this embodiment, the magnetic particle brake 8 serves as the core component for torque control, enabling adaptive torque compensation. It adjusts the output torque in real time according to the instructions from the control module to adapt to the user's training state. A DC magnetic particle brake can be used; this is not a limitation.

[0070] The output shaft of the magnetic powder brake is connected to one end of the rotating shaft 26 via a coupling to ensure stable torque transmission without impact, while compensating for the coaxiality deviation between the rotating shaft and the magnetic powder brake.

[0071] The rotary encoder 16 is used to detect the rotation state of the gear 27 in real time, and then calculates the movement speed and push-off displacement of the slide plate 5, providing supplementary data support for the control module to adjust the torque. An incremental rotary encoder can be used, and there is no limitation here.

[0072] The input shaft of the rotary encoder is connected to the other end of the rotating shaft 26 via a coupling, and rotates synchronously with the gear 27 and the rotating shaft 26. By detecting the rotation angle of the rotating shaft and combining it with the transmission ratio of the gear and rack (number of gear teeth / pitch of rack teeth), the linear displacement of the rack (i.e. the push-off displacement of the skateboard) can be calculated. At the same time, the speed of the skateboard can be calculated by the rate of change of the rotation angle.

[0073] This embodiment adds a magnetic powder brake, which has a fast response speed and a wide torque adjustment range, so as to adjust the output torque in real time and accurately. Combined with the data from the attitude sensor and rotary encoder, the torque compensation is more in line with the user's training state, avoiding training ineffectiveness or damage caused by torque deviation, and further improving the training's relevance and reliability.

[0074] The rotary encoder can detect the rotation state of the gears in real time, accurately calculate the speed of the skateboard and the displacement of the push-off, and combine it with posture data to fully understand the user's training status (posture, force speed, force amplitude), providing more comprehensive data support for precise torque adjustment. At the same time, it can provide real-time feedback of training data, making it convenient for users to understand their training progress.

[0075] In some embodiments, the training device further includes a touch control module 2 for displaying a preset gluteal muscle training mode.

[0076] The touch control module 2 can be mounted on the top edge of the housing 1 using a fixed bracket.

[0077] During use, the touch control module 2 clearly displays a variety of preset glute training modes using an icon + text display method. Users can quickly identify and select the mode that suits their needs without complicated operations, solving the problems of cumbersome mode selection and lack of intuitive display in existing training devices. It is especially suitable for novice users, lowering the operating threshold. The touch control module 2 also includes a speaker for sound output.

[0078] Thirdly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements any of the above-mentioned adaptive torque compensation methods for gluteal muscle training based on pelvic compensation feedback.

[0079] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0080] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0081] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device embodiments are described simply because they are similar to the method embodiments; relevant parts can be referred to the descriptions of the method embodiments.

[0082] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes to the above embodiments within the scope of the present invention without departing from the principles and spirit of the present invention.

Claims

1. A method for adaptive torque compensation in gluteal muscle training based on pelvic compensatory feedback, characterized in that, The method includes: The real-time push-off displacement during the concentric abduction phase of unilateral glute training is obtained, and the real-time displacement ratio is calculated by combining it with the pre-stored maximum push-off length of the user. The lever arm compensation coefficient is determined based on the real-time displacement ratio. Obtain the baseline average velocity of the first movement and the current average velocity of the current movement during the concentric abduction phase, and calculate the velocity retention rate; The speed fatigue adjustment coefficient is adjusted stepwise according to the speed retention rate. The system obtains the initial pelvic pitch angle when the user is in an upright and neutral position, as well as the current pelvic tilt angle and current pitch angle during the training process, and determines the pitch angle change based on the current pitch angle and the initial pitch angle. If the current pelvic tilt angle does not exceed the preset safe tilt range and the pitch angle change does not exceed the preset safe pitch change threshold, the attitude penalty coefficient is set to 0; otherwise, the attitude penalty coefficient is calculated using the following two-dimensional proportional feedback model. : ； in, For roll penalty gain, For pitch penalty gain, As the maximum penalty limit, This represents the current pelvic tilt angle. This represents the change in pitch angle. To pre-set a safe roll range, To preset a safe pitch change threshold; Using speed fatigue adjustment coefficient Attitude penalty coefficient And lever arm compensation coefficient The initial resistance is preset based on the current training mode. F Dynamic adjustments are made to obtain the target damping force for the next action. T To perform dynamic resistance control, the updated model is as follows: 。 2. The method according to claim 1, characterized in that, The maximum push-off length is obtained through the following steps: Monitor a single concentric stretching motion of the lower limbs under baseline damping force; When the change in pelvic pitch angle exceeds the preset safe pitch change threshold, the push-off displacement at this time is obtained as the maximum push-off length. The step of determining the lever arm compensation coefficient based on the real-time displacement ratio includes: when At that time, the lever arm compensation coefficient ; when At that time, the lever arm compensation coefficient ; when At that time, the lever arm compensation coefficient ; in The preset peak safety factor, This is the upper limit of the initial power zone. This represents the lower limit of the peak contraction zone. This represents the real-time displacement ratio.

3. The method according to claim 1, characterized in that, The stepwise adjustment of the speed fatigue adjustment coefficient based on the speed retention rate includes: when At that time, speed fatigue adjustment coefficient ; when At that time, speed fatigue adjustment coefficient ; when At that time, speed fatigue adjustment coefficient ; in, R For speed retention rate, R 1 represents the first fatigue threshold. R 2 represents the second fatigue threshold. For the attenuation slope, This represents the minimum safety factor.

4. The method according to claim 1, characterized in that, The method further includes: During continuous training of the same gluteal muscles, record the number of repetitions N and determine the number of repetitions. The average speed of movement during the concentric abduction phase of this movement Calculate the average value of the action sequence. and the overall average speed The fatigue descent slope of the current training side was calculated using the least squares method. : ； ； ； The gluteal muscles on the currently trained side are determined based on the vector direction of the current pelvic tilt angle, combined with the fatigue descent slope. K Identify the weaker gluteal muscle and automatically increase the target damping force during the eccentric phase for the weaker gluteal muscle in subsequent training.

5. The method according to claim 4, characterized in that, The automatic increase of the target damping force for the weaker gluteal muscle during the eccentric phase in subsequent training includes: During the centripetal extension phase where v > 0, the target damping force is: T During the centrifugal recovery phase when v < 0, the target damping force is... T Multiply by centrifugal overload factor O ecc As the actual output damping force, among which O ecc >1.

6. The method according to claim 1, characterized in that, The method further includes: The damping force output by the equipment during the period of peak contraction [ t 1, t Integrating within 2], we obtain the peak tension impulse for a single instance used to assess the depth of gluteal muscle stimulation. I : ; Among them, the push-off displacement is not less than c× M The dynamic range is the peak contraction region, and c is a preset constant proportional coefficient. M This represents the user's maximum push-off length. Output the glute mass control index after training. G Its calculation model is as follows: ； ； in, N The number of times the action is completed. For the first The effective range of motion for each movement. The target tension impulse preset by the system, For the first The posture penalty coefficient triggered by the next action. , , All are preset weighting coefficients. x This is for real-time pedal displacement.

7. An adaptive torque-compensated gluteal muscle training device, characterized in that, The training device performs the method described in any one of claims 1-6 during use; The training device includes: a housing (1), an attitude sensing acquisition device (3), a flexible pad (4), a sliding plate (5), a rack (18), a slide rail (23), a rotating shaft (26), and a gear (27); The rotating shaft (26) is rotatably connected inside the housing (1), and the rotating shaft (26) is connected to the gear (27). One end of the rack (18) is located inside the housing (1) and meshes with the gear (27). The other end of the rack (18) is located outside the housing (1), and the other end of the rack (18) is connected to the slide plate (5). The slide plate (5) is slidably connected to the slide rail (23). The flexible pad (4) is disposed on the slide plate (5). The attitude sensing acquisition device (3) is adapted to be worn on the user's pelvic area and is configured to collect pelvic attitude data in real time. The pelvic attitude data includes tilt angle, pitch angle and lateral acceleration.

8. The training device according to claim 7, characterized in that, The training device also includes: a skateboard base (6), a first bracket (7), a magnetic powder brake (8), a second bracket (15), a rotary encoder (16), a front rack retainer (17), a slide (21), a slider (22), a crank (24), and a rear rack retainer (25). The first bracket (7) and the second bracket (15) are both disposed inside the housing (1). The magnetic powder brake (8) is disposed on the first bracket (7), and the rotary encoder (16) is disposed on the second bracket (15). The rotating shaft (26) is rotatably connected between the first bracket (7) and the second bracket (15). One end of the rotating shaft (26) is connected to the magnetic powder brake (8), and the other end is connected to the rotary encoder (16). The rotary encoder (16) is configured to detect the rotation of the gear (27) in real time to calculate the movement speed and pedal displacement. The portion of the rack (18) located inside the housing (1) is supported by the front rack retainer (17) and the rear rack retainer (25); the end of the rack (18) located outside the housing (1) is provided with a crank (24) and a slider (22), the slider (22) is set in the groove (21) on the slide rail (23), and the slider (22) is connected to the slide plate (5) and the slide plate base (6).

9. The training device according to claim 7, characterized in that, The training device also includes a touch control module (2) for displaying preset gluteal muscle training modes.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1-6.

Images