Knee joint exoskeleton motion recognition method and knee joint power assisting device

Abstract

The invention relates to the field of exoskeleton power assisting devices, in particular to a knee joint exoskeleton motion recognition method and a knee joint power assisting device.The knee joint exoskeleton motion recognition method comprises a lower limb exoskeleton device, the lower limb exoskeleton device comprises a joint driving motor, and the joint driving motor comprises a pair of rotating assemblies; the joint driving motor is used for applying torque to the pair of rotating assemblies rotating relatively, and the relative rotating center of the rotating assemblies corresponds to the knee joint; the thigh extending frame and the shank extending frame are connected with the pair of rotating assemblies respectively; the thigh binding and locking device is mounted on the thigh extension frame and is bound on the thigh; the first shank binding and locking device and the second shank binding and locking device are installed on the shank extending frame and bound between the calf and the knee joint. The anti-falling device has the advantages of good anti-falling effect and good use adaptability. And during downhill movement, knees can be protected through control of simulated damping.

Description

Technical Field

[0001] The present invention relates to the field of exoskeleton assistive devices, and particularly to a method for recognizing the movement of a knee exoskeleton and a knee assistive device. Background Art

[0002] The knee exoskeleton device binds the user's thigh and calf and uses a motor located at the knee to drive the lower limb to perform an assistive movement.

[0003] Existing lower limb knee exoskeletons are difficult to accurately recognize the movement state of the exoskeleton only relying on sensors located on the lower limb. This requires an additional position detection sensor outside the lower limb exoskeleton, resulting in cumbersome wearing, or extending the lower limb exoskeleton above the waist and hip and installing the position detection sensor at the position of the lower limb exoskeleton corresponding to above the waist and hip, which makes the lower limb knee exoskeleton too heavy. Summary of the Invention

[0004] To overcome the above-mentioned deficiencies of the prior art, improve the recognition accuracy of the movement state, and be convenient to wear while ensuring the lightweight of the device.

[0005] To achieve the above object, the present invention is implemented by the following technical solutions: A method for recognizing the movement of a knee exoskeleton, based on a lower limb exoskeleton device, the lower limb exoskeleton device includes a thigh extension frame and a calf extension frame respectively strapped to the thigh and calf, the thigh extension frame and the calf extension frame are relatively rotatably arranged around the knee joint, and a movement parameter sensor is provided on the lower limb exoskeleton device; the method includes the following processes: Collecting the movement parameters of the lower limb exoskeleton device through the movement parameter sensor to determine the movement characteristics, and then confirming the movement state according to the movement characteristics; The movement characteristics include: knee joint angle, knee joint angular velocity, inclination angle of the calf and / or thigh, angular velocity of the calf and / or thigh, acceleration value of the thigh and / or calf, and hip projection difference, and the hip projection difference is the change amount of the hip position in the current state relative to the preset initial posture of the lower limb.

[0006] A knee assistive device, including a lower limb exoskeleton device, the lower limb exoskeleton device includes: A joint drive motor, the joint drive motor includes a pair of rotating components, the joint drive motor is used to apply torque to a pair of relatively rotating rotating components, and the center of relative rotation of the rotating components corresponds to the knee joint; A thigh extension frame and a calf extension frame respectively connected to a pair of rotating components, the thigh extension frame and the calf extension frame are respectively strapped to the thigh and calf.

[0007] It can be seen from the above technical solutions that the present invention has the following beneficial effects: The knee joint angle, angular velocity, thigh / calf tilt angle and acceleration value are directly collected through motion parameter sensors (such as IMU, encoder) on the lower limb exoskeleton, and combined with the hip projection difference (the change in lower limb spatial displacement calculated based on geometry), to achieve a comprehensive recognition of the motion state.

[0008] The hip projection difference is calculated by the trigonometric function relationship between the lower limb length and joint angle, replacing the function of the traditional waist and hip position sensor, avoiding the device from extending upward to the waist and hip, and significantly simplifying the structure.

[0009] The sensors are all integrated on the thigh extension frame and calf extension frame of the lower limb exoskeleton, without the need for additional waist and hip components to be tied, reducing the wearing steps and the weight of the device. Description of the Drawings

[0010] Figure 1 Schematic diagram of the use of a knee joint assistance device in an embodiment of the present application; Figure 2 Structural diagram of the front of a knee joint assistance device in an embodiment of the present application (hiding the binding parts); Figure 3 Structural diagram of the back of a knee joint assistance device in an embodiment of the present application (hiding the binding parts); Figure 4 Schematic diagram of the principle corresponding to the hip projection difference in an embodiment of the present application.

[0011] In the figure: 1 - joint drive motor; 2 - thigh extension frame; 21 - thigh locking device; 211 - upper connection seat; 212 - first binding part; 3 - calf extension frame; 31 - first calf locking device; 311 - lower connection seat; 312 - second binding part; 32 - second calf locking device; 321 - knee connection seat; 322 - third binding part. Detailed Description of the Invention

[0012] Embodiment 1 Combined with Figures 1 to 3 As shown, this embodiment provides a knee joint assistance device, including a lower limb exoskeleton device, and the lower limb exoskeleton device includes: Joints drive motor 1, the joints drive motor 1 includes: a pair of rotating components, the joints drive motor 1 is used to apply torque to a pair of relatively rotating components, and the center of the relative rotation of the rotating components corresponds to the knee joint; a thigh extension frame 2 and a calf extension frame 3 respectively connected to the pair of rotating components; a thigh locking device 21 installed on the thigh extension frame 2, and the thigh locking device 21 is bound to the thigh; specifically, in this embodiment, the pair of rotating components are the stator and rotor of the joints drive motor 1. It also includes a motion parameter sensor installed on the lower limb exoskeleton device, and the motion parameter sensor is used to collect the motion parameters of the lower limb exoskeleton device to determine the motion characteristics.

[0013] When the existing knee joint exoskeleton is bound to the lower limb, it is easy to sag, and the binding position needs to be adjusted according to users with different leg lengths, which has the disadvantage of poor use adaptability. In this regard, further, a knee joint assisting device in this embodiment also includes a first calf locking device 31 and a second calf locking device 32 installed on the calf extension frame 3, and the first calf locking device 31 and the second calf locking device 32 are bound between the calf belly and the knee joint.

[0014] Based on the above structure, the principle of a knee joint assisting device is: the joints drive motor 1 is used to apply torque to the rotating components, and the force is transmitted to the thigh and calf through the thigh extension frame 2 and the calf extension frame 3 to assist the movement of the knee joint. In this application, the first calf locking device 31 and the second calf locking device 32 are arranged above the calf belly and below the knee joint. The thigh locking device 21, the first calf locking device 31 and the second calf locking device 32 form three anchor points for connecting the lower limb exoskeleton device to the lower limb. In this embodiment, on the one hand, the protruding characteristic of the calf belly has the effect of preventing sagging. On the other hand, the distance between the calf belly and the knee joint in the lower limb part is relatively short. By placing the two anchor points of the calf here, it can adapt to users with different leg lengths and has the advantage of good use adaptability. In an embodiment, in order to further prevent the lower limb exoskeleton device from sagging, the lower limb exoskeleton device can be bound to the waist or above the waist.

[0015] Further, in this embodiment, the thigh locking device 21 includes an upper connecting seat 211 and a first binding member 212. The upper connecting seat 211 is rotatably installed on the thigh extension frame 2, the rotation axis of the upper connecting seat 211 is horizontal, the upper connecting seat 211 fits against the back of the thigh, and the first binding member 212 binds the thigh to the upper connecting seat 211.

[0016] Further, in this embodiment, the first lower leg locking device 31 includes a lower connecting seat 311 and a second binding member 312. The lower connecting seat 311 is rotatably mounted on the lower leg extension frame 3. The rotation axis of the lower connecting seat 311 is horizontal. The lower connecting seat 311 fits against the back side of the lower leg. The second binding member 312 binds the lower leg to the lower connecting seat 311. The rotatably arranged lower connecting seat 311 and the upper connecting seat 211 have better conformity with the lower limbs.

[0017] Further, in this embodiment, the second lower leg locking device 32 includes a knee connecting seat 321 and a third binding member 322. The knee connecting seat 321 is mounted on the lower leg extension frame 3. The knee connecting seat 321 fits against the side of the lower leg. The third binding member 322 binds the lower leg to the knee connecting seat 321. The knee connecting seat 321 is located above the lower connecting seat 311. Specifically, the binding member is a binding strap and is connected to the connecting seat by a magnetic buckle.

[0018] Further, in this embodiment, the joint drive motor 1 is arranged beside the knee joint. The thigh extension frame 2 and the lower leg extension frame 3 are integrally bent plate members. The ends of the thigh extension frame 2 and the lower leg extension frame 3 away from the joint drive motor 1 extend to the back side of the leg. The thigh locking device 21 and the first lower leg locking device 31 are respectively installed at the ends of the thigh extension frame 2 and the lower leg extension frame 3 away from the joint drive motor 1. It has the advantages of reasonable structure and light weight.

[0019] Further, in this embodiment, it further includes a power module. The power module is electrically connected to the joint drive motor 1. The power module is placed at the waist or above the waist.

[0020] Further, in this embodiment, it further includes motion parameter sensors. The number of motion parameter sensors is 2, which are used to collect the motion parameters of the lower limb exoskeleton device to determine at least one motion feature among the knee joint angle, the inclination angles of the lower leg and the thigh, the angular velocity, and the hip projection difference. It minimizes the use of motion parameter sensors and has the advantage of low cost.

[0021] In one embodiment, the motion parameter sensors are a pair of IMUs respectively installed on the thigh extension frame 2 and the lower leg extension frame 3. The knee joint angle can be obtained by converting the motion data such as the inclination angle and the angular velocity obtained by the pair of IMUs.

[0022] In another embodiment, the motion parameter sensor is an IMU mounted on the thigh extension 2 or the calf extension 3 and a knee joint angle sensor. The motion states (such as angular velocity and acceleration) corresponding to the extension without the IMU can be calculated from the motion data obtained by the IMU and the change in the knee joint angle obtained by the knee joint angle sensor. In one embodiment, the knee joint angle sensor is an encoder provided on the joint drive motor 1. The encoder is used to monitor the angular change between the stator and the rotor on the joint drive motor 1, so as to obtain the change signals of the knee joint angle and the angular velocity.

[0023] Embodiment 2 Based on a knee joint assistance device provided in Embodiment 1, this embodiment provides a control method, including the following steps: Motion parameter sensor data processing, obtaining the detection data input by the motion parameter sensor provided on the lower limb exoskeleton, and processing the detection data to extract motion features; Confirm the motion state according to the motion features.

[0024] Specifically, the motion states include gait phase and activity type.

[0025] Further, in one embodiment, the control method further includes: Gait phase detection, determining the current gait phase according to the motion features. The gait phase includes the stance phase and the swing phase, where the stance phase corresponds to the state of the lower limb touching the ground, and the swing phase corresponds to the state of the lower limb leaving the ground; Activity recognition, extracting motion features in the stance phase to determine the activity type, and the activity type includes at least one of walking, going up, going down, running, squatting, walking backward, and unknown; Control strategy, selecting a corresponding control strategy according to the gait phase and the activity type to control the joint drive motor 1.

[0026] Among them, the stance support phase (hereinafter referred to as the "stance phase") is defined as the gait phase when the lower limb is in contact with the ground, which usually marks the beginning of a complete gait cycle. The occurrence of this stage is universal. Regardless of the activity type of the wearer (such as walking, going up, squatting or going down), the gait always starts from the contact of the foot with the ground. Therefore, the stance phase provides a natural time window that can be used as the starting point of the gait cycle to capture and analyze the initial features of the gait.

[0027] Moreover, in the stance phase, the motion parameter sensor data (such as the acceleration from the thigh inertial measurement unit IMU and the knee joint angular velocity) shows a relatively stable trend. For example, the acceleration approaches the acceleration due to gravity (about 9.8 m / s²), and the angular velocity amplitude is small (usually below a certain threshold, such as 0.5 rad / s). This stability provides a reliable reference signal for feature extraction, in contrast to the fluctuations in the subsequent dynamic phases (such as the swing phase).

[0028] On the one hand, the extraction of motion features depends on the key characterization of the motion pattern. Due to its specific biomechanical and kinetic properties, the stance phase becomes the best time point for identifying the activity type. Specifically manifested as follows: 1. In the stance phase, the body posture of the wearer (such as knee joint angle, calf tilt angle, hip position, etc.) directly reflects the type of activity. For example, when squatting, the knee joint angle decreases significantly (e.g., less than 120 degrees), and when going uphill, the calf tilt angle is larger (e.g., more than 30 degrees). These posture information are most obvious and stable when the foot contacts the ground, facilitating capture through motion parameter sensor data. Therefore, in contrast, during the swing phase, the foot leaves the ground, and the posture changes are more driven by inertia, making it difficult to directly correlate to a specific activity type.

[0029] 2. The motion parameter sensor signals in the stance phase have a lower noise level because the contact between the foot and the ground reduces the uncertainty brought by free movement. This stability allows the algorithm to more accurately extract features, such as the smoothness of acceleration, the low amplitude of knee joint angular velocity, or the static value of knee joint angle. While in the swing phase, the signal fluctuations are larger (for example, the acceleration may vary by 10% - 50% due to the swing speed), and the features are easily interfered by environmental factors (such as uneven ground) or individual differences (such as leg swing habits).

[0030] 3. As the starting point of a gait cycle, the stance phase provides the earliest opportunity to identify the activity type. This early identification is crucial for the subsequent detection and control of gait phases. For example, after identifying the "going uphill" activity, the algorithm can adjust the switching conditions of the swing phase (such as increasing the knee joint angular velocity threshold), thereby optimizing the phase determination of the entire gait cycle.

[0031] Furthermore, in one embodiment, the motion features include knee joint angle, knee joint angular velocity, tilt angle of the calf and / or thigh, angular velocity of the calf and / or thigh, and acceleration of the thigh and / or calf.

[0032] Further, in one embodiment, the motion feature includes the hip projection difference, which is the change in the hip position of the current state relative to the hip position when the lower limbs are in a preset initial posture. Specifically, it refers to the difference in the position change of the hip when moving in the horizontal direction (front and back) and the vertical direction (up and down), and is used to capture the overall trend of leg movement. Simply put, it "projects" the movement of the hip onto the ground and the sagittal plane, including the amount of forward and backward movement (X direction) and the amount of up and down movement (Y direction), and then calculates the difference between the two. This difference can characterize whether the wearer's leg tends to step forward, step backward, or be lifted or lowered more. For example, when walking forward, the projection in the horizontal direction is usually larger than the change in the vertical direction; when walking backward, the horizontal direction may become smaller or even negative, while the vertical direction may increase slightly. By measuring the hip projection difference, different motion states can be distinguished more clearly, such as the difference between normal walking and walking backward. Specifically, as shown in Figure 3 When the lower limbs are in the preset initial posture, the thigh and the calf are in an upright state, that is, the knee joint angle is 180°. The hip projection difference includes the hip projection difference in the X direction and the hip projection difference in the Y direction; The hip projection difference in the X direction can be expressed as the vector sum of the horizontal displacements. The positive direction in the X direction is horizontal forward: ΔX = -[l_shk · cos(θ_shk)+ l_thg · cos(θ_thg)] where: ΔX represents the hip projection difference of the hip in the X direction (unit: meter); l_shk and l_thg are the lengths of the calf and the thigh respectively (typical values are both 0.42 meters); θ_shk and θ_thg are the angles of the calf and the thigh relative to the vertical direction respectively (unit: radian); The hip projection difference in the Y direction is combined with normalization processing. The positive direction in the Y direction is vertically upward and can be expressed as: ΔY = l_shk · sin(θ_shk) + l_thg· sin(π - θ_thg) - L_total where: ΔY represents the hip projection difference of the hip in the Y direction (unit: meter); l_shk and l_thg are the lengths of the calf and the thigh respectively (typical values are both 0.42 meters); θ_shk and θ_thg are the angles of the calf and the thigh relative to the vertical direction respectively (unit: radian); sin(π - θ_thg); L_total = l_shk + l_thg is the total length of the leg and is used for normalization. After normalization processing, it is applicable to users of different heights or leg lengths.

[0033] The motion features based on the hip projection difference also include the change rate of the hip projection difference, that is, calculating the time derivative of the hip projection difference, including the differential of the hip projection difference in the X direction of the hip joint and the differential of the hip projection difference in the Y direction of the hip joint, and then capturing the change rate of the hip projection difference through a smoothing differentiator to further enhance the description of the motion trend.

[0034] In one embodiment, the hip projection difference plays a role in gait phase detection in the following ways: 1. Switch from the stance phase to the swing phase: When switching from the stance phase to the swing phase, the hip projection difference provides the spatial context of the body center of gravity and the leg position. For example, when the hip projection difference in the X direction becomes negative (e.g., less than -0.12 m) and the hip projection difference in the Y direction shows an upward offset (e.g., greater than -0.08 m), combined with the knee joint angular velocity exceeding 1.5 rad / s, the system determines that the leg begins to leave the ground and enters the swing phase.

[0035] The dynamic change of the hip projection difference (the first-order differential with respect to time) further assists in judging the switching timing. For example, when the first-order differential of the hip projection difference on the hip x-axis > 0 and the first-order differential of the hip projection difference on the hip y-axis < -0.1, it indicates that the hip moves forward and the leg is lifted upward, strengthening the identification of the swing phase.

[0036] 2. Switch from the swing phase to the stance phase: When switching from the swing phase to the stance phase, the hip projection difference is used to detect the spatial motion features of foot contact with the ground. For example, when the hip projection difference on the hip y-axis reaches a positive peak (e.g., greater than 0.15 m) and its first-order differential value (with respect to time) > 0.25, it indicates that the leg has completed the swing and re-contacted the ground, triggering the determination of the stance phase.

[0037] The extreme value of the hip projection difference combined with a time window (e.g., within 150 milliseconds after the peak occurs) ensures the robustness of the switching condition.

[0038] 3. Enhancement of the activity type context: The hip projection difference is not only used for phase switching, but also provides support for activity type recognition through its trend and amplitude. For example, when going uphill, the hip projection difference in the X direction is usually small (e.g., less than 0.05 m) while the hip projection difference in the Y direction shows a significant negative value (e.g., less than -0.07 m), reflecting the spatial characteristics of the leg-lifting action. These motion features are used to dynamically adjust the switching threshold to adapt the phase detection to specific activities.

[0039] Compared with other motion features, the hip projection difference has the following unique advantages: 1. Comprehensiveness of spatial context: The hip projection difference directly reflects the two-dimensional displacement (in the x and y directions) of the leg relative to the body's center of gravity, providing more comprehensive kinematic space information than single joint angles or angular velocities. For example, knee joint angles only describe the local joint state, while the hip projection difference captures the overall coordination between the hip and the leg. This global perspective is particularly important in complex activities such as going up or down stairs, as these activities involve significant shifts in the body's center of gravity.

[0040] 2. Sensitivity to center of gravity movement: The hip projection difference is highly sensitive to changes in the wearer's center of gravity and can effectively distinguish the stability of the stance phase from the dynamics of the swing phase. For example, during walking, the process of the hip projection on the x-axis changing from negative to positive reflects the trend of the center of gravity moving from the rear to the front, while the negative change in the hip projection on the y-axis indicates the height of leg lift. This biaxial kinematic feature is superior to relying solely on knee joint angular velocity, which may lose its discriminative ability due to individual gait differences.

[0041] 3. Robustness and environmental adaptability: The hip projection difference is based on geometric calculations (trigonometric functions) and is less sensitive to sensor noise (such as acceleration drift in IMUs). In contrast, acceleration kinematic features are easily affected by external disturbances (such as uneven ground), while the hip projection difference maintains high stability through normalized calculations of leg length. This robustness gives it an advantage in dynamic and variable scenarios such as mountain climbing.

[0042] 4. Implementation with low computational overhead: The calculation of the hip projection difference only involves basic trigonometric operations and addition and subtraction, with low time complexity and no need for complex filtering or iterative optimization. Compared with combined kinematic features based on acceleration and angular velocity (such as deviation ratio or curvature change), the hip projection difference provides sufficient discriminative ability while maintaining high efficiency, meeting the design goal of the lightweight algorithm of the present invention.

[0043] If the hip projection difference is removed and only other kinematic features are relied on for gait phase detection, the following adverse consequences may occur: 1. Decreased accuracy of phase switching: Without the spatial context of the hip projection difference, the system may have difficulty accurately determining the position of the leg relative to the body. For example, during the activity of going downstairs, if only angular velocity is relied on and the positive peak of the hip projection on the x-axis is ignored, a short leg adjustment may be misjudged as the swing phase, resulting in incorrect application of control strategies (such as incorrect application of transparent control during the stance phase).

[0044] 2. Weakened ability to distinguish activity types: The hip projection difference provides crucial spatial motion features in activity recognition, such as the negative trend of the hip projection difference in the Y direction during uphill movement. Without this motion feature, the system may not be able to effectively distinguish between walking and going uphill (the angular velocity patterns of both may be similar), which affects the accuracy of dynamic threshold adjustment and ultimately leads to inappropriate assistive output.

[0045] 3. Reduced robustness: In complex terrains (such as uneven ground during mountain climbing), angular velocity and acceleration are vulnerable to noise interference, while the hip projection difference maintains relative stability through geometric relationships. Without using the hip projection difference, the system's adaptability to environmental changes will significantly decrease, and it may frequently enter unknown phases under abnormal conditions (such as brief distortion of sensor signals), reducing the safety and comfort of the wearer.

[0046] 4. Impaired transition smoothness: The rate of change of the hip projection difference (the first derivative with respect to time) provides a smooth dynamic basis for gait phase switching. Without this motion feature, the switching conditions may overly rely on instantaneous values (such as angular velocity thresholds), resulting in jitter or sudden changes in gait phase transition, affecting the continuity of exoskeleton control and the user experience.

[0047] Therefore, the hip projection difference (in the x and y directions) significantly improves the accuracy and adaptability of gait phase detection by providing spatial context, center-of-gravity sensitivity, and robustness. Its unique advantage lies in efficiently representing complex motion patterns at low computational cost, enabling the system to maintain stable phase switching and control output during various activities (such as mountain climbing, going uphill). Without using the hip projection difference, the system will face problems such as decreased accuracy, weakened adaptability, and uneven transitions, limiting its performance in dynamic scenarios.

[0048] In one embodiment, the hip projection difference plays a role in activity type judgment in the following ways: The hip projection differences in the X direction and Y direction of the hip are spatial displacement indicators calculated based on the angles and lengths of the calf and thigh, used to characterize the position changes of the knee exoskeleton during movement. By combining the hip projection difference and its derivative (i.e., the rate of change of the hip projection difference with respect to time), the activity type of the user can be effectively judged, such as going downstairs, going upstairs, or walking. These hip projection differences reflect the relative motion trajectories of the hip in the horizontal (X direction) and vertical (Y direction), which are closely related to leg postures and movement dynamics.

[0049] In the determination of activity types, the hip projection difference in the X direction is mainly used to detect the displacement motion characteristics in the horizontal direction. For example, in the activity of going downstairs, the hip projection difference in the X direction usually shows a small positive or negative offset. Combining its differential can further capture the dynamic changes of the motion, so as to distinguish the patterns of going downstairs quickly and slowly. In the activity of walking, the hip projection difference in the X direction usually shows a periodic negative offset, reflecting the motion characteristics of the legs moving forward. In the activity of going upstairs, the hip projection difference in the X direction may show a large positive offset, indicating the action of lifting and pushing forward the legs.

[0050] The hip projection difference in the Y direction mainly reflects the displacement changes in the vertical direction and is often used to detect activities related to height changes. For example, when going downstairs, the hip projection difference in the Y direction is usually negative, indicating that the hip moves downward relative to the initial position; when going upstairs, the hip projection difference in the Y direction may show a change from negative to positive first, reflecting the process of lifting the leg and landing. By normalizing the hip projection difference in the Y direction (subtracting the total leg length), the influence of individual leg length differences can be eliminated, making the judgment result more universal.

[0051] By integrating the hip projection differences in the X and Y directions and their differentials, the system can construct a multi-dimensional motion feature space, which, combined with other sensor data (such as knee joint angles, calf and thigh angles, gyroscope data, etc.), forms the discriminant conditions for activity types. This method can capture the spatio-temporal characteristics of the motion, thus realizing the accurate classification of different activity types.

[0052] In this regard, the hip projection difference has the following unique advantages: 1. Spatial intuitiveness: The hip projection difference directly reflects the displacement of the hip in three-dimensional space, providing more intuitive motion trajectory information than a single angle or speed, which helps to distinguish activities with similar angular motion characteristics but different spatial paths (such as walking and going downstairs).

[0053] 2. Dynamic adaptability: By introducing the differential of the hip projection difference, the acceleration and trend changes of the motion can be captured, making the system more sensitive to the detection of dynamic activities (such as sudden leg lifting when going downstairs quickly or upstairs).

[0054] 3. Robustness: The hip projection difference combines the length and angle information of the calf and thigh, and can offset the influence of sensor noise or individual posture differences to a certain extent, being more stable than directly using the original sensor data.

[0055] 4. Multi-dimensional cooperation: The combination of the hip projection differences in the X and Y directions provides a motion description on a two-dimensional plane, which is complementary to other motion characteristics (such as the angular velocity of the knee joint), enhancing the accuracy and specificity of activity classification.

[0056] If the hip projection differences in the X and Y directions of the hip are not used, the activity type judgment may face the following problems: 1. Reduced resolution ability: Relying solely on a single motion feature such as angle or angular velocity may not be able to effectively distinguish activities with similar spatial trajectories but different dynamics. For example, going downstairs and walking may have a similar range of knee joint angles, but the horizontal and vertical displacement patterns of the hip are significantly different. Removing the hip projection differences will lead to an increase in the misjudgment rate.

[0057] 2. Lack of dynamic information: The differential of the hip projection difference provides real-time change information on the motion trend. Without using it, the system may have difficulty capturing rapid movements or transition states (such as switching from standing to going downstairs), thus affecting the response speed and stability of the control algorithm.

[0058] 3. Reduced adaptability: The normalized design of the hip projection difference makes it applicable to users with different heights or leg lengths. Without using the hip projection difference, the system may require additional calibration steps to adapt to individual differences, increasing the complexity of use.

[0059] 4. Limited control accuracy: In the control of the knee exoskeleton, the hip projection difference provides a key basis for gait phase switching and torque distribution. Without this information, the control algorithm may not be able to accurately match the user's motion intention, resulting in insufficient or excessive assistance from the exoskeleton, affecting comfort and safety.

[0060] In summary, the hip projection difference plays an irreplaceable role in activity type judgment and can significantly improve the control performance of the knee exoskeleton in complex terrains and dynamic activities.

[0061] Furthermore, in one embodiment, the motion feature includes the product of the thigh acceleration and the knee joint angular velocity to characterize the mechanical interaction during motion. Specifically, the product of the thigh acceleration and the knee joint angular velocity is designed as an intermediate variable. The core idea is to extract more discriminative features by fusing two different dimensions of motion parameters (linear acceleration and angular velocity) to enhance the accuracy of activity type recognition and gait phase detection. Specific applications in some embodiments include the following aspects: 1. Characterization of mechanical interaction This feature directly reflects the dynamic interaction between the linear motion of the thigh and the rotational motion of the knee joint. For example, during the upward movement, when the wearer raises the leg, the thigh acceleration may present a positive peak (e.g., 12 m / s²), and at the same time, the knee joint angular velocity increases rapidly (e.g., 2.0 rad / s), and the product can reach 24 m·rad / s³, indicating a strong mechanical coupling. During flat walking, due to the lower knee joint angular velocity (e.g., 0.8 rad / s), even if the acceleration is similar, the product value is smaller (e.g., 9.6 m·rad / s³). This difference provides a significant basis for activity classification.

[0062] 2. Enhancement of Dynamic Trend Through multiplication, this feature amplifies the dynamic changes during the movement process. For example, when quickly transitioning from the swing phase to the stance phase (such as when running and landing), the thigh acceleration may rapidly decrease from a high value (e.g., from 15 m / s² to 9.8 m / s²), while the knee joint angular velocity changes from a positive value to nearly zero (e.g., from 1.5 rad / s to 0.2 rad / s). The rapid change in the product (from 22.5 m·rad / s³ to 1.96 m·rad / s³) can sensitively capture this trend, thus assisting in the precise determination of gait phase switching.

[0063] 3. Robustness of Noise Suppression When using thigh acceleration or knee joint angular velocity alone, the signal may be affected by environmental noise (such as acceleration jitter caused by uneven ground) or sensor drift. By multiplying the two, this engineering feature smooths the instantaneous anomalies of a single variable to a certain extent. For example, if the acceleration shows an abnormal peak due to a short-term interference (such as 20 m / s²), but the knee joint angular velocity remains at a low value (such as 0.3 rad / s), the product value (6 m·rad / s³) still remains within a reasonable range, avoiding misjudgment.

[0064] Specific applications in activity recognition: When extracting motion features during the stance phase to determine the activity type, the product of thigh acceleration and knee joint angular velocity is used as a key discriminant index. In some embodiments, for example: Walking on flat ground: It is manifested as the product value showing periodic fluctuations, typically in the range of 5 to 15 m·rad / s³, reflecting a stable gait rhythm.

[0065] Going uphill: It is manifested as the product value increasing significantly when lifting the leg (e.g., 20 to 30 m·rad / s³) because both the acceleration and angular velocity reach their peaks simultaneously.

[0066] Squatting down: It is manifested as the product value being relatively small and lasting for a long time (e.g., 2 to 5 m·rad / s³, lasting for about 500 milliseconds), reflecting the characteristics of low speed and high load.

[0067] Going downhill: It is manifested as the product value being medium and changing smoothly (e.g., 10 to 20 m·rad / s³) because the angular velocity is low and the acceleration tends to be stable.

[0068] By setting threshold rules and combining time window analysis (such as the average value within 300 milliseconds), the system can efficiently distinguish different activity types.

[0069] Role in Control Strategy The product of thigh acceleration and knee joint angular velocity also directly affects the selection of actuator control strategies. For example: When it is higher than a certain threshold (such as 25 m·rad / s³), it indicates that the wearer is in high dynamic activities (such as running or going uphill), and the system can select hybrid control to provide strong assistance by combining elastic and damping effects.

[0070] When it is lower and stable (such as 3 to 8 m·rad / s³), it indicates that the wearer is in a quasi-static state (such as squatting or standing), and the system can adopt elastic control to simulate the spring effect.

[0071] Compared with using only thigh acceleration or knee joint angular velocity, the product of thigh acceleration and knee joint angular velocity has the following advantages: 1. Information fusion: Integrate linear and rotational motion information in a product form to provide a richer mechanical description than a single variable.

[0072] 2. Dynamic sensitivity: The product operation amplifies the co-variation of the two variables, making the feature more sensitive to the response of the motion intention.

[0073] 3. Computational efficiency: Only one multiplication operation is required to generate the feature, which meets the lightweight algorithm design goal and is suitable for real-time processing of embedded systems.

[0074] 4. Strong adaptability: The range and trend of the product value can adapt to the motion habits of different wearers (such as differences in stride or speed) without additional calibration.

[0075] If the product of thigh acceleration and knee joint angular velocity is removed and only the original variables are relied on, the following problems may occur: Reduced discrimination ability: It is difficult for a single variable to capture the coupling effect of linear and rotational motions simultaneously, and it may confuse dynamically similar activities (such as walking flat and going uphill slowly).

[0076] Insufficient dynamic detection: Without the amplification effect of the product feature on the change trend, the system may miss the critical time points of rapid transitions (such as swing to support).

[0077] Imprecise control: The control strategy may not be able to accurately match the motion requirements due to the lack of comprehensive mechanical indicators, resulting in insufficient or excessive assistance.

[0078] Furthermore, in one embodiment, the motion feature includes the normalized thigh acceleration - knee joint angular velocity ratio. By applying an upper threshold to limit the influence of outliers, the robustness of the feature is enhanced.

[0079] The core of the normalized thigh acceleration - knee joint angular velocity ratio lies in the positive or negative nature of the knee joint angular velocity, which reflects whether the knee joint is in a flexed state (e.g., the angular velocity is positive when lifting the leg) or an extended state (e.g., the angular velocity is negative when lowering the leg). By distinguishing this directionality, the system can better understand the intention of the wearer's leg movement. For example, during the swing phase, the knee joint usually flexes rapidly, while during the stance phase, the knee joint may extend slowly or remain stable.

[0080] On this basis, the relationship between thigh acceleration and knee joint angular velocity is further combined. Specifically, it generates a ratio by comparing the linear acceleration of the thigh with the rotational speed of the knee joint. This ratio is not simply obtained by dividing the two, but is normalized to ensure that its value is within a controllable range, avoiding inconsistencies caused by differences in the magnitudes of the original data (e.g., acceleration is usually in meters per second squared, while angular velocity is in radians per second). The normalized ratio can more intuitively reflect the relative strength of the two. For example, during rapid leg lifting, both thigh acceleration and knee joint angular velocity may be relatively high, while during slow squatting, both may be relatively low. In addition, an upper limit threshold is introduced to limit extreme values caused by sensor noise, external interference, or abnormal movements of the wearer. For example, when the wearer is suddenly impacted by an external force, the thigh acceleration may show a brief abnormal peak while the knee joint angular velocity remains normal, and the upper limit threshold can limit the ratio within a reasonable range to avoid misjudgment by the system.

[0081] This feature performs excellently in various application scenarios of lower limb exoskeletons, especially suitable for activity type recognition and gait phase detection. For example: Gait analysis during level walking: During normal level walking, the knee joint angular velocity is positive (knee joint flexion) during the swing phase, and approaches zero or is negative (knee joint extension) during the stance phase. The normalized ratio can highlight the synergistic effect of the rapid increase in thigh acceleration and angular velocity during the swing phase, helping the system accurately distinguish between the swing and stance phases.

[0082] Dynamic capture during ascending: During ascending, the knee joint angular velocity shows a significant positive value during the leg - lifting phase, and at the same time, the thigh acceleration increases significantly due to the leg lift. The normalized ratio reflects this high - dynamic state and avoids abnormal fluctuations caused by uneven ground or excessive force exerted by the wearer through the upper limit threshold.

[0083] Stability judgment during squatting or standing: During squatting, the knee joint angular velocity is usually negative (knee joint slowly extends), and the thigh acceleration is also low and stable. The value of the normalized ratio is small and changes smoothly, indicating a low - speed motion state, and the system can adjust to a low - assistance mode accordingly.

[0084] Filtering of abnormal movements: When the wearer's thigh acceleration suddenly surges due to a fall or external impact, the knee joint angular velocity may not change synchronously, and the normalization ratio may show an abnormally high value. After the upper limit threshold intervenes, the ratio is restricted within a reasonable range to prevent the system from misidentifying it as a certain high-dynamic activity.

[0085] The reasons for enhancing its robustness include: Positive and negative nature of the knee joint angular velocity: By focusing on the directionality of the angular velocity, the feature can naturally distinguish different stages of movement and avoid the ambiguity of a single numerical value. For example, even if the absolute value of the angular velocity is the same, the positive and negative differences can clearly distinguish the flexion and extension intentions.

[0086] Normalization processing: Normalization ensures that the ratio is not directly affected by the dimension of sensor data or individual differences of the wearer (such as weight, stride). For example, a heavier wearer may generate greater acceleration, but after normalization, the ratio still follows the same trend as that of a lighter wearer, enhancing the generality of the feature.

[0087] Protective effect of the upper limit threshold: The upper limit threshold is like a safety valve, avoiding the interference of abnormal values on the system's judgment. For example, during running, if the acceleration soars instantaneously due to a protrusion on the ground, the threshold mechanism can prevent the ratio from getting out of control and ensure that the system still operates in the normal running mode.

[0088] The ratio of the externally normalized thigh acceleration to the knee joint angular velocity can support the optimization of the control strategy for lower limb exoskeletons. In some embodiments, for example: When the ratio is high and the knee joint angular velocity is positive, it indicates that the wearer is in a state of quickly lifting the leg (such as running or going uphill), and the system can increase the assistive output of the actuator to provide greater thrust.

[0089] When the ratio is low and the angular velocity is negative, it indicates that the wearer may be slowly lowering the leg (such as squatting or going downstairs), and the system can switch to the damping mode to enhance stability and reduce impact.

[0090] When the ratio triggers the upper limit threshold due to an abnormal value, the system can temporarily maintain the current control strategy, avoiding frequent mode switching due to short-term interference, thereby improving the coherence of the wearing experience.

[0091] Furthermore, in one embodiment, the motion feature includes accumulating the positive knee joint angular velocity to quantify the continuous trend of knee joint movement; when the knee joint angular velocity is negative, the accumulated value is reset.

[0092] Furthermore, in one embodiment, the motion feature includes the maximum / minimum values of the angles and angular velocities of the knee joint, thigh, and calf, characterizing the peak trend of the movement.

[0093] Further, in one embodiment, the control strategy includes at least one of elastic control, damping control, and hybrid control; where: Elastic control is to control the torque of the joint drive motor 1 acting on the swing arm 11 according to the knee joint angle to simulate a spring; Damping control is to control the torque of the joint drive motor 1 acting on the swing arm 11 according to the angular velocity of the knee joint to simulate a damper; Hybrid control is to combine elastic and damping control to adapt to the current activity type; The control strategy corresponding to the stance phase includes at least one of elastic control, damping control, and hybrid control.

[0094] When executing the control strategy, the control parameters will be adjusted in real time according to the current gait phase and / or activity type and / or movement characteristics. Such as adjusting the elastic force of the simulated spring or the damping force of the simulated damper.

[0095] Regarding the control strategy corresponding to the standing support phase, it should be noted that: During the standing support phase (hereinafter referred to as the "support phase"), the system detects the motion state when the foot contacts the ground and applies specific control strategies to optimize the stability, comfort, and efficiency of the wearer. These strategies include elastic control, damping control, and hybrid control, each designed for different motion requirements and scenarios.

[0096] For elastic control, the system applies torque according to the joint angle to simulate the behavior of a virtual spring. For example, when the wearer is in a semi-squat posture (knee joint angle is about 90 to 120 degrees, usually occurring when squatting from a standing position), the system detects this angle change, indicating that the leg is in a static load or quasi-static transition state. At this time, the system applies an auxiliary torque proportional to the angle offset (for example, the torque magnitude is 5 to 15 Nm, specifically depending on the wearer's weight and joint stiffness requirements) to reduce the burden on the leg muscles and simulate the natural spring-back effect. The benefit of this control is that it not only enhances the wearer's strength by providing assistance but also reduces energy consumption during posture recovery (such as standing up from a squat), while avoiding discomfort caused by excessive joint bending. The joint angle is detected by an inertial measurement unit (IMU), which combines accelerometer and gyroscope data to calculate the relative angle of the leg in real time, ensuring detection accuracy and response speed.

[0097] For damping control, the system applies torque according to the angular velocity of the knee joint to simulate the buffering effect of a damper. For example, when the wearer walks on uneven ground or squats quickly, the angular velocity of the knee joint may reach 0.5 to 1.5 rad / s, indicating that the leg is undergoing dynamic adjustment or external disturbances (such as ground impact). In this scenario, the system applies a reverse torque proportional to the angular velocity of the knee joint (e.g., 3 to 10 Nm, depending on the magnitude of the angular velocity of the knee joint) to slow down the joint movement speed and avoid instability or muscle fatigue caused by excessive bending. The advantage of this control is that it can smooth the movement trajectory and suppress unnecessary oscillations (such as jitters during walking), thereby enhancing the safety and comfort of the wearer.

[0098] Hybrid control combines elastic and damping effects and dynamically adjusts according to the type of activity. For example, in the scenario of "carrying a heavy object and squatting", the wearer may first enter a semi-squat (triggering elastic control to provide 10 Nm of spring torque assistance), and then due to the shaking of the heavy object, rapid joint adjustment occurs (the angular velocity of the knee joint reaches 1.0 rad / s, triggering damping control to apply 5 Nm of buffering torque). This combined strategy ensures that the system adaptively matches the needs of the wearer in complex activities by providing both assistance and stability at the same time. Its advantage lies in integrating the force enhancement of elasticity and the movement smoothness of damping, and being able to adapt to diverse stance phase scenarios (such as ascending, squatting and rising, or walking with a load), thereby enhancing the overall movement efficiency and wearing experience.

[0099] The fundamental reason for adopting these control strategies during the stance phase is that this gait phase is a crucial point for activity type recognition and load support. The actions of the wearer (such as walking, squatting, or standing adjustment) usually start from the stance phase. At this time, the joint angle and the angular velocity of the knee joint provide rich motion information, enabling the system to accurately judge the activity intention and optimize the torque output.

[0100] In addition, the stability and assistance requirements during the stance phase directly affect the smoothness of the subsequent swing phase. Therefore, through the synergistic effect of elastic, damping, and hybrid control, the system not only improves the performance of the current gait but also lays the foundation for phase transition.

[0101] Furthermore, in one embodiment, the control strategy includes at least one of transparent control and drag reduction control; wherein: The control strategy corresponding to the swing phase includes transparent control and / or drag reduction control.

[0102] Regarding the control strategy corresponding to the swing phase, it should be noted that: During the swing phase, the foot leaves the ground and the leg is in a free movement state. The system aims to minimize interference or provide mild enhancement to support natural movement and improve efficiency. The control strategy includes transparent control and drag reduction control, respectively targeting different dynamic requirements.

[0103] Transparent control allows the wearer to move the leg in a natural way by setting the torque output to zero. For example, during normal walking, when the leg enters the swing phase (such as the process from heel-off to toe-touching the ground), the knee joint swings freely at an angular velocity of about 0.8 to 1.2 rad / s. At this time, the system does not apply any active torque, and only relies on the low-friction design of the device (such as achieved through an efficient motor and transmission system) to ensure that the wearer does not perceive additional resistance. The advantage of this control is that it maximally preserves the wearer's natural gait, avoids discomfort or energy waste caused by unnecessary intervention, and at the same time maintains the high transparency of the device, making it suitable for daily activities (such as walking on flat ground or jogging).

[0104] Drag reduction control, on the other hand, counteracts the effects of friction or gravity by applying a small assistive torque. For example, during long-distance walking or climbing, the knee joint in the swing phase may increase the wearer's burden due to the self-weight of the device (usually 2 to 5 kg) or terrain resistance. At this time, the system detects a decrease in the angular velocity of the knee joint (such as below 0.5 rad / s, indicating that the swing is blocked), and applies a mild assistive torque (for example, 1 to 3 Nm, in the direction of the swing) to compensate for the resistance caused by device friction and gravity. The benefit of this control is that it reduces the wearer's muscle fatigue without changing the natural movement trajectory, especially significantly improving efficiency during long-term activities or high-intensity tasks (such as mountain climbing).

[0105] The reason for selecting these control strategies during the swing phase is that the core requirements of this gait phase are freedom of movement and efficiency. Transparent control meets the requirements of natural movement through zero torque output, while drag reduction control optimizes the wearer's experience by fine-tuning the torque to make up for the inherent physical limitations of the device (such as mass or friction). In addition, the control during the swing phase directly affects the coherence of the gait cycle, and lightweight intervention strategies can ensure seamless connection with the stance phase, avoiding gait disorders caused by excessive control.

[0106] Furthermore, in one embodiment, the gait phase also includes an unknown phase, which corresponds to at least one of the following situations: the transitional state when the motion characteristics do not clearly conform to the stance or swing phase, the algorithm cannot determine the current gait, or during algorithm initialization; the control strategy for the unknown phase includes transparent control and / or drag reduction control to ensure safety and comfort.

[0107] Regarding the control strategy corresponding to the unknown phase, it should be noted that: When the system cannot clearly determine the current gait phase (i.e., in the unknown phase), transparent control or drag reduction control is preferentially adopted to ensure safety and comfort. This strategy is applicable to scenarios where the gait transition is ambiguous or the motion parameter sensor signals are uncertain.

[0108] In the unknown phase, the system defaults to transparent control and maintains the torque output at zero. For example, when the wearer suddenly starts walking from a standing position, but the initial motion parameter sensor data (such as IMU signals) cannot accurately distinguish between support and swing due to noise or brief occlusion, the system avoids applying any active torque and only relies on the passive characteristics of the device to support movement. The advantage of this method is that it avoids the risk of misjudgment (such as applying damping torque during the swing phase and causing tripping) through zero intervention, thus ensuring the safety and smoothness of the wearer's movements.

[0109] If a slight motion trend is detected (such as the knee joint angular velocity between 0.2 and 0.5 rad / s, indicating a possible entry into the swing or fine-tuning of the posture), the system can switch to resistance reduction control and apply a small torque (for example, 0.5 to 2 Nm) to support the potential movement intention. For example, when the wearer slowly adjusts their standing posture or prepares to take a step, this slight assistance can offset the device friction and ensure the smooth start of the movement. The advantage of this strategy is that it can still provide appropriate support under uncertainty while minimizing intervention and avoiding disturbing the wearer.

[0110] The reasons for adopting these control strategies in the unknown phase are safety and robustness. When the gait phase is not clear, any aggressive control (such as high torque output) may cause discomfort or even danger. Therefore, giving priority to transparent control can effectively avoid risks. The resistance reduction control serves as a conservative supplement to ensure that the system remains responsive under possible movement intentions. This design not only improves the fault tolerance of the system but also enables it to maintain reliability and wearer comfort in complex or non-standard scenarios (such as irregular gaits or sudden movements).

[0111] Furthermore, in one embodiment, it also includes correspondingly adjusting the gait transition conditions and / or the control parameters corresponding to the control strategies according to the activity type and / or motion characteristics, ensuring that the exoskeleton adapts to changing scenarios.

[0112] The motion characteristics extracted in combination with the support phase are not only used to classify the current activity but also directly affect the logic of subsequent gait switching: specifically manifested as: I. The transition from the support phase to the swing phase. According to the activity type identified in the support phase, the algorithm dynamically adjusts the switching conditions. For example, if it is identified as "squatting", a longer acceleration stabilization time (such as 500 milliseconds) is required to confirm the end of the support phase to avoid misjudging the start of the swing due to a brief rise; if it is identified as "going up", the stabilization time may be shortened (such as 150 milliseconds) and the knee joint angular velocity threshold may be increased (such as 1.5 rad / s) to adapt to rapid leg raising. This adjustment ensures the accuracy of gait phase determination and the relevance of gait switching.

[0113] II. Implementation of Seamless Control: By extracting motion features in the support phase, the system can establish a motion context in the early stage of the gait cycle, thus providing continuous instructions for the control of exoskeletons or walking aids. For example, the features identified during the "downward" activity can trigger a more gentle actuator response and avoid sudden changes.

[0114] It should be noted that: Phase transition refers to the process in which the exoskeleton system switches from one gait phase (such as the stance support phase) to another gait phase (such as the swing phase or unknown phase) according to the wearer's real-time motion state. This transition is achieved by detecting specific gait features (such as joint angles, knee joint angular velocity) and comparing them with predefined thresholds or conditions. The core logic of gait phase transition relies on the fusion of multi-motion parameter sensor data and dynamic threshold judgment to ensure that the system can accurately identify the wearer's motion intention and timely adjust the control strategy.

[0115] In the stance support phase, the system continuously monitors the filtered value of the knee joint angular velocity. When this parameter breaks through the preset dynamic threshold (the typical value is 0.8 rad / s) and the hip forward projection offset reaches 0.25 m, the transition to the swing phase is triggered. This process adopts a triple verification mechanism: first, the continuity of the motion trend is captured by the thigh inertial measurement unit, second, the joint angle extreme value locking technology is used to suppress instantaneous noise interference, and finally, the confidence check is carried out in combination with the phase prediction model of historical gait data to ensure the reliability of the transition determination.

[0116] The adjustment of the transition logic by activity type is reflected in two aspects: threshold dynamic compensation and control domain remapping. Taking the upward scenario as an example, the system increases the knee joint angular velocity determination threshold to 1.5 rad / s to compensate for the response delay caused by the increased joint load, and at the same time introduces a kinematic compensation algorithm for the hip forward projection offset to dynamically correct the forward movement of the body center of gravity during stair climbing.

[0117] Furthermore, in one embodiment, the motion features also include the frequency domain features of knee joint motion. In complex terrain scenarios, the system characterizes the ground roughness through the frequency domain features of knee joint motion (such as the tremor component of 2 to 4 Hz), and dynamically couples the elastic coefficient of the simulated spring in the elastic control and the damping coefficient of the simulated damper in the damping control. For example, when the terrain complexity is too large, that is, the ground roughness exceeds the critical value, the damping coefficient is proportionally increased to the maximum value, and at the same time, the elastic coefficient is attenuated to 60% of the reference value, effectively suppressing the joint oscillation caused by irregular impacts.

[0118] For low-speed and high-load activities such as squatting and standing up, the system constructs a non-linear elastic control strategy: when the knee joint angle enters the 60°-100° working range, an auxiliary torque that is exponentially related to the angle change rate is applied, and this torque generates a peak assist (typical value 12 Nm) when the angle reaches 90°. During this process, the hip projection difference in the Y direction is used as a support phase maintenance condition. When it exceeds 0.1 m for 200 ms, the system will forcibly lock the current control parameters to prevent mis-switching caused by accidental center of gravity fluctuations.

[0119] The abnormal working condition handling mechanism adopts a simple safety strategy: when entering an unknown phase, the system automatically switches to the transparent control mode.

[0120] In some embodiments, the methods of gait phase detection and gait phase switching condition control are specifically as follows: Before adjusting the switching conditions according to the detected activity type, the gait phase switching conditions are based on predefined rules and characteristic trends of sensor data, specifically as follows: The switching condition from the swing phase to the support phase is: when the sensor detects that the foot touches the ground, for example, the thigh acceleration value tends to be stable (for example, close to the typical range of gravitational acceleration, such as near 9.8 m / s²), and the magnitude of the knee joint angular velocity decreases significantly (for example, below a certain preset threshold, such as 0.5 rad / s), it is determined to enter the support phase. Because, stable acceleration and low knee joint angular velocity reflect the static characteristics of the movement after the foot touches the ground, in contrast to the dynamic changes in the swing phase.

[0121] The switching condition from the support phase to the swing phase is: when the sensor detects that the foot leaves the ground, for example, the thigh acceleration shows significant fluctuations (such as in the range of ±10%), and the magnitude of the knee joint angular velocity increases (for example, exceeds a certain preset threshold, such as 1.0 rad / s), it is determined to enter the swing phase. Because, the acceleration fluctuations and the increase in knee joint angular velocity indicate that the foot begins to leave the ground and enters a dynamic motion state.

[0122] The condition for determining an unknown gait phase is: when the sensor data does not clearly meet the characteristic conditions of the support phase or the swing phase, for example, the combination of the measured acceleration at the thigh and the knee joint angular velocity is in the fuzzy interval of the predefined threshold (for example, the knee joint angular velocity fluctuates between 0.5 - 1.0 rad / s and the acceleration is not stable), or the data is affected by noise interference resulting in unclear characteristics, it is determined to be an unknown gait phase. As the default processing of the algorithm for transitional states or uncertain states, it usually occurs at the boundary of gait phase switching or during algorithm initialization.

[0123] These unadjusted switching conditions are often based on fixed thresholds and direct trend analysis of sensor data, and are applicable to general gait patterns (such as walking on flat ground).

[0124] Furthermore, after the activity type changes, the gait phase switching conditions will be adjusted. That is, when a specific activity (such as going up, going down, squatting, etc.) is detected, the gait phase switching conditions will be dynamically adjusted according to the activity characteristics to improve the detection accuracy and adaptability. In some embodiments, the specific adjustment methods are as follows: Characterize the context of the current movement through the activity characteristics (such as knee joint angle, calf tilt angle, hip projection difference, etc.) extracted by the activity recognition module.

[0125] Adjust the threshold or feature weight according to the activity type. For example, a higher knee joint angular velocity threshold may be required when going up, while a longer duration requirement for the stance phase may be needed when squatting.

[0126] For example, in some embodiments, the switching conditions when going up are as follows: The switching condition from swing to stance: The acceleration stabilization time measured at the thigh is required to be shorter (for example, shortened from 200 milliseconds to 150 milliseconds), and at the same time, the peak value of the knee joint angular velocity needs to be higher (for example, adjusted from 1.0 rad / s to 1.5 rad / s) to adapt to the faster foot touchdown.

[0127] The switching condition from stance to swing: Detecting a larger knee joint angle change (for example, exceeding 60 degrees) as an additional condition to reflect the leg-lifting action.

[0128] For example, in some embodiments, the switching conditions when going down are as follows: The switching condition from swing to stance: Lower the knee joint angular velocity threshold (for example, adjusted from 1.0 rad / s to 0.8 rad / s) to adapt to the slower falling action.

[0129] The switching condition from stance to swing: Extend the acceleration fluctuation detection window (for example, extended from 200 milliseconds to 300 milliseconds) to capture a smooth transition.

[0130] For example, in some embodiments, the switching conditions when squatting are as follows: Prolong the stance phase: Require that the time when the acceleration is stable and the knee joint angle significantly decreases (for example, less than 120 degrees) exceeds a certain threshold (such as 500 milliseconds) to avoid misjudgment as a short pause.

[0131] By dynamically adjusting the threshold and time window, the switching conditions can better match the motion characteristics of specific activities and avoid misjudgment. For example, prevent misjudgment as the swing phase due to fast touchdown when going up, or prevent misjudgment as the start of the swing due to a long stance time when squatting.

[0132] Furthermore, to ensure seamless control of gait phase switching, the following control logic is adopted: Smoothing transition mechanism: Use overlapping time windows (e.g., 50 - 100 milliseconds) at the gait phase transition boundary to smooth the motion features. For example, reduce the switching jitter caused by noise or transient fluctuations by performing a moving average (such as 5 - point or 10 - point average) on the thigh acceleration and knee joint angular velocity.

[0133] When a potential transition is detected (e.g., from stance to swing), the algorithm briefly maintains the current gait phase state (e.g., for an additional 50 milliseconds) before confirming the new gait phase to verify the persistence of the feature trend.

[0134] State continuity check: Require that the features meet the switching conditions for a minimum duration (e.g., 100 milliseconds) to avoid false switching triggered by short - term anomalies (such as sensor jitter). For example, when switching from the stance phase to the swing phase, confirm the switch only when the knee joint angular velocity continuously exceeds the threshold and the acceleration fluctuations persist.

[0135] Dynamic threshold update: Adjust the threshold in real - time according to the activity type to ensure that the switching conditions are consistent with the current motion context. For example, increase the angular velocity threshold during running to match the faster swing frequency. If the activity type changes (e.g., from walking to ascending), the algorithm updates the threshold within the first complete gait cycle after detecting the new activity to avoid misjudgment of the gait phase during the transition period.

[0136] Fallback mechanism: If an anomaly is detected after the switch (e.g., the acceleration suddenly stabilizes after switching to the swing phase), the algorithm can fallback to the previous gait phase within a short time (e.g., within 100 milliseconds), mark it as an unknown gait phase, and wait for the data to stabilize before re - judging. This mechanism is especially applicable to transitional states or situations where sensor signals are unstable.

[0137] Results of seamless control: Through the above logic, this method ensures the continuity and accuracy of gait phase switching and avoids interruption of control signals. For example, in the application of exoskeletons or walking aids, the smoothness of gait phase switching can be directly translated into seamless responses of motor control, improving the comfort and safety of the wearer.

[0138] Furthermore, in one embodiment, the process of activity recognition includes: Feature comparison based on threshold rules: Use predefined threshold rules to compare the motion features and determine whether the conditions for a specific activity are met; Extreme value detection within a time window, which analyzes the extreme values (such as maximum or minimum) of motion characteristics using a time window (e.g., 100 to 350 milliseconds) to capture the dynamic changes of motion. For example, by detecting the maximum value of the calf tilt angle and the peak of the hip projection difference, it determines whether there are characteristic patterns related to ascending or descending, and then filters through time conditions (e.g., the duration after the peak occurs) to ensure the robustness of the detection; Cumulative counting and state persistence check, which cumulatively counts the motion characteristics that meet the preset conditions, such as the number of times the calf angle peak is continuously detected exceeding a certain threshold, or the duration for which the knee joint angular velocity remains within a certain range. Through state persistence check (e.g., requiring a combination of motion characteristics to continuously meet the conditions for more than a certain time), it avoids misjudgment caused by short-term noise or abnormal data; Dynamic update, within the stance phase, dynamically updates the motion characteristics, such as calculating the product of the knee joint angular velocity and the hip projection difference, or the feature integration based on the time window, to characterize the trend and intensity of motion. Through these dynamically updated motion characteristics, different activity types are distinguished; for example, ascending may exhibit greater knee joint angle changes and negative hip projection differences, while descending may exhibit more stable angular velocities of the knee joint and thigh and a specific calf angle range.

[0139] Conditional logic combination, uses conditional logic combinations (such as "AND", "OR" logic) to jointly analyze multiple motion characteristics. For example, if the knee joint angle exceeds a certain threshold, and at the same time the knee joint angular velocity shows a positive trend, and the hip projection difference is less than a certain negative value, it may be determined as an ascending activity. This logical combination avoids complex calculation models and relies only on basic mathematical operations (such as addition, subtraction, multiplication, division, comparison), thus achieving lightweight.

[0140] Therefore, in this application, the process of activity recognition is a lightweight analysis method based on the trend of motion parameter sensor data, which does not rely on resource-intensive technologies (such as machine learning or deep learning), but is based on direct feature extraction and threshold logic judgment of real-time motion parameter sensor data. Through predefined rules and time window analysis, the computational complexity is reduced to ensure efficient operation on embedded devices. All feature calculations use low-overhead mathematical operations (such as addition, multiplication, comparison) and combine with cached data within a finite time window to avoid large-scale data storage and processing.

[0141] Through the above lightweight analysis method, various activity types can be efficiently distinguished. For example: Level walking: It is characterized by stable fluctuations in knee joint angular velocity and changes in knee joint angle.

[0142] Ascending: It is characterized by larger knee joint angle changes, negative hip projection differences in the Y direction, and specific calf tilt angles.

[0143] Downward: Characterized by a stable knee joint angular velocity and a small calf tilt angle.

[0144] Running: Characterized by high-frequency fluctuations in knee joint angular velocity and a significant increase in thigh acceleration. Specifically, during running, the knee joint angular velocity shows rapid and periodic positive and negative alternations, reflecting the rapid flexion of the leg during the swing phase (lifting the leg) and the brief extension during the support phase (landing). At the same time, the thigh acceleration reaches its peak during the swing phase, especially showing a large positive change when the leg moves forward, and a short deceleration impact when landing. The knee joint angle changes significantly and rhythmically, while the calf tilt angle switches frequently with each step, presenting an overall highly dynamic motion pattern.

[0145] Squatting: Characterized by a significant decrease in knee joint angle and a longer support phase duration.

[0146] Backward walking: Characterized by the hip projection difference in the X direction tending to be negative and the hip projection difference in the Y direction tending to be positive, and the cumulative value of the knee joint angular velocity showing a positive trend.

[0147] Unknown: When the characteristic pattern does not conform to any known activity conditions, it is classified as unknown.

[0148] The technical principles of the present invention have been described above in combination with specific embodiments. These descriptions are only for explaining the principles of the present invention and cannot be construed in any way as a limitation on the protection scope of the present invention. Based on the explanations herein, those skilled in the art can readily conceive of other specific embodiments of the present invention without creative efforts, and these embodiments will fall within the protection scope of the present invention.

Claims

1. A method for knee exoskeleton motion recognition, characterized in that: Based on a lower limb exoskeleton device, the lower limb exoskeleton device includes a thigh extension frame (2) and a calf extension frame (3) respectively strapped to the thigh and the calf. The thigh extension frame (2) and the calf extension frame (3) are relatively rotatably arranged around the knee joint. A motion parameter sensor is provided on the lower limb exoskeleton device. The following process is included: The motion parameter sensor is used to collect the motion parameters of the lower limb exoskeleton device to determine the motion characteristics, and then the motion state is confirmed according to the motion characteristics; The motion characteristics include: knee joint angle, knee joint angular velocity, inclination angle of the calf and / or the thigh, angular velocity of the calf and / or the thigh, acceleration value of the thigh and / or the calf, and hip projection difference. The hip projection difference is the change amount of the hip position in the current state relative to the hip position when the lower limb is in a preset initial posture.

2. A knee joint assisting device, characterized in that: Including a lower limb exoskeleton device, the lower limb exoskeleton device includes: A joint drive motor (1), the joint drive motor (1) includes a pair of rotating components, the joint drive motor (1) is used to apply torque to a pair of relatively rotating rotating components, and the center of relative rotation of the rotating components corresponds to the knee joint; The thigh extension frame (2) and the calf extension frame (3) respectively connected to a pair of rotating components, the thigh extension frame (2) and the calf extension frame (3) are respectively strapped to the thigh and the calf; It further includes a motion parameter sensor installed on the lower limb exoskeleton device. The motion parameter sensor is used to collect the motion parameters of the lower limb exoskeleton device to determine the motion characteristics. The motion characteristics include: knee joint angle, knee joint angular velocity, inclination angle of the calf and / or the thigh, angular velocity of the calf and / or the thigh, acceleration value of the thigh and / or the calf, and hip projection difference. The hip projection difference is the change amount of the hip position in the current state relative to the hip position when the lower limb is in a preset initial posture.

3. The knee joint assist device according to claim 2, characterized in that: The motion parameter sensor is a pair of IMUs respectively installed on the thigh extension frame (2) and the calf extension frame (3).

4. The knee joint assisting device according to claim 2, characterized in that: The motion parameter sensor is an IMU installed on the thigh extension frame (2) or the calf extension frame (3) and a knee joint angle sensor.

5. The knee joint assist device according to claim 2, wherein, It further includes: A thigh locking device (21) installed on the thigh extension frame (2), and the thigh locking device (21) is strapped to the thigh; A first calf locking device (31) and a second calf locking device (32) installed on the calf extension frame (3), and the first calf locking device (31) and the second calf locking device (32) are strapped between the calf belly and the knee joint.

6. The knee joint assist device according to claim 5, wherein: The thigh locking device (21) includes an upper connecting seat (211) and a first binding member (212). The upper connecting seat (211) is rotatably installed on the thigh extension frame (2). The rotation axis of the upper connecting seat (211) is horizontal. The upper connecting seat (211) fits against the back of the thigh, and the first binding member (212) straps the thigh to the upper connecting seat (211).

7. The knee joint assist device according to claim 5, wherein: The first lower leg locking device (31) includes a lower connecting seat (311) and a second binding member (312). The lower connecting seat (311) is rotatably mounted on the lower leg extension frame (3). The rotation axis of the lower connecting seat (311) is horizontal. The lower connecting seat (311) fits against the back side of the lower leg. The second binding member (312) binds the lower leg to the lower connecting seat (311).

8. The knee joint assisting device according to claim 5, wherein: The second lower leg locking device (32) includes a knee connecting seat (321) and a third binding member (322). The knee connecting seat (321) is mounted on the lower leg extension frame (3). The knee connecting seat (321) fits against the side of the lower leg. The third binding member (322) binds the lower leg to the knee connecting seat (321). The knee connecting seat (321) is located above the lower connecting seat (311).

9. The knee joint assistance device according to claim 5, wherein: The joint drive motor (1) is arranged beside the knee joint. The thigh extension frame (2) and the lower leg extension frame (3) are integrally curved plate members. The ends of the thigh extension frame (2) and the lower leg extension frame (3) far from the joint drive motor (1) extend to the back side of the leg. The thigh locking device (21) and the first lower leg locking device (31) are respectively mounted on the ends of the thigh extension frame (2) and the lower leg extension frame (3) far from the joint drive motor (1).

10. A knee joint assist device according to claim 2, characterized in that: It further includes a power supply module. The power supply module is electrically connected to the joint drive motor (1). The power supply module is placed at the waist or above the waist.

Images