Active power wearable assisting equipment

Abstract

The invention discloses active power wearable assistance equipment, and aims to solve the problems of bulkiness, short endurance, inaccurate recognition, poor man-machine coupling and the like of existing equipment so as to realize the application targets of light weight, long endurance, intelligent and accurate assistance, safety, reliability and comfort in wearing. The active power wearable assistance equipment adopts a mechanical structure design conforming to ergonomics, and comprises a chest controller, a back main supporting frame, a waist protection plate, a leg supporting frame, a wearable shoe cover and an assistance driving structure. The chest controller senses the movement intention of a wearer in real time through a sensor system, the electric push rods on the thighs and the shanks are controlled to stretch out and draw back synergistically through a hierarchical algorithm, accurate and smooth active assistance is provided in key stages of actions such as walking, stooping and standing up, and the equipment has good wearing comfort, adaptability and adjustability and is suitable for being used in the field of wearable equipment. The device can be widely applied to power line repair, material carrying, rescue after disasters, high-altitude operation and other scenes, and has remarkable practical value and popularization prospect.

Description

Technical Field

[0001] This invention relates to a wearable assistive device, specifically a wearable assistive device based on an exoskeleton structure with active driving capability, which is particularly suitable for high-intensity work scenarios such as power maintenance, emergency rescue, and material handling. Background Technology

[0002] In fields such as power, fire fighting, emergency rescue, and logistics handling, workers often need to perform heavy-duty work in high-intensity, complex terrain, and even dangerous environments, such as power equipment repair, long-distance material transportation, and post-disaster site disposal. Such work not only places extremely high demands on the physical fitness of personnel, but also easily leads to chronic strain on the lower back and knee joints due to repetitive bending, standing, climbing, and long-term heavy-duty movements. In addition, work fatigue can even cause secondary safety accidents such as falls from heights or being struck by objects.

[0003] To alleviate the workload of workers, wearable assistive devices have emerged. Currently, these devices are mainly divided into passive and active categories. Passive devices typically use elastic elements to store and release energy. While simple in structure, they suffer from limited assist direction and effectiveness, and generate significant resistance when bending or squatting, interfering with natural movements. Actively powered devices provide assistance through motors and other active drive units, offering greater performance potential. However, current products still have significant drawbacks. For example, the HAL exoskeleton from Japan's Cyberdyne, while capable of accurate intention recognition through bioelectrical signals, is excessively heavy (approximately 23 kg) and has a short battery life (less than 3 hours), severely limiting its application in long-duration, large-scale operations. Similar products developed domestically, such as those by the State Grid Cixi Power Supply Company, suffer from insufficient lumbar support due to their split design, poor human-body coupling, and inaccurate electromyography signal recognition leading to delayed assist response.

[0004] In summary, existing wearable assistive devices generally suffer from one or more technical bottlenecks, such as bulky structure, insufficient battery life, inaccurate motion intention recognition, and poor coupling with the human body. These limitations make it difficult to put them into practical application in typical power emergency repair scenarios such as mountainous areas, high altitudes, and confined spaces, and they cannot effectively meet the urgent needs of high-intensity operations to reduce personnel load, ensure operational safety, and improve efficiency. Summary of the Invention

[0005] To address the aforementioned shortcomings of existing technologies, this paper provides an active-powered wearable assistive device, aiming to solve the problems of bulkiness, short battery life, inaccurate recognition, and poor human-machine coupling in existing devices, so as to achieve the application goals of lightweight, long battery life, intelligent and precise assistance, safety and reliability, and comfortable wear.

[0006] The technical solution adopted in this invention is as follows: An active-powered wearable assistive device includes a chest controller, a back support frame, a lumbar support, a leg support frame, wearable shoe covers, and an assistive drive structure. The front chest controller integrates a main controller, power supply, and signal processing module. It is connected to the main back support frame via a set of shoulder straps and a set of chest safety straps. The main back support frame conforms to the curve of the human back, and a set of lumbar support panels is fixedly installed at its bottom. The set of lumbar support panels conforms to the curves of the sides of the human waist and is connected to two sets of leg support frames in the downward direction. A lumbar safety strap is installed between the sets of lumbar support panels on the front side of the human waist. Shoe covers are installed at the bottom of the two sets of leg support frames. The power assist drive structure includes a first electric push rod mounted on the thigh exoskeleton arm and a second electric push rod mounted between the lower leg exoskeleton arm and the wearing shoe cover. The first and second electric push rods are electrically connected to the front chest controller. The front chest controller senses the wearer's movement status and intention in real time through a sensor system integrated throughout the equipment, and coordinates the output force, speed, and stroke of the first and second electric push rods to provide timely and appropriate auxiliary torque during key stages of movements such as walking, bending over, and standing up, thereby completing the power assist.

[0007] Furthermore, the front-chest controller adopts an insulating encapsulation structure to insulate and encapsulate the main controller, power supply, and signal processing module; The main back support frame has a U-shaped structure, which includes a support back plate and a set of back frames fixedly connected to the support back plate; the support back plate is connected to the front chest controller through a set of chest safety straps, which are equipped with buckles; the set of back frames is connected to the front chest controller through a set of shoulder straps.

[0008] Furthermore, the lumbar support plate is hinged to the back support plate on the back of the human body, and its lower end extends to the outside of the sides of the human thighs and is hinged to the leg support frame. A load-bearing platform is also installed between the main back support frame and a set of lumbar support plates. The load-bearing platform is a flat plate that is located behind the human body and is hinged to the back support plate and a set of back frames. A spring preload adjustable connector is hinged between the load-bearing platform and the set of lumbar support plates.

[0009] Furthermore, the leg support frame includes a thigh exoskeleton arm and a calf exoskeleton arm; one end of the thigh exoskeleton arm is hinged to the lumbar support plate, and the other end is hinged to one end of the calf exoskeleton arm, and the other end of the calf exoskeleton arm is hinged to the shoe cover; the thigh exoskeleton arm is equipped with a thigh elastic strap that is fixed to the human thigh. The first electric push rod of the power-assisted drive structure is hinged at one end to the upper end of the thigh exoskeleton arm and at the other end to the upper end of the calf exoskeleton arm; the second electric push rod is hinged at one end to the middle and lower section of the calf exoskeleton arm and at the other end to the heel of the shoe cover. The shoe covers are equipped with adjustable buckles for easy wearing.

[0010] Furthermore, the sensor system includes an inertial measurement unit (IMU) for acquiring leg angular velocity and acceleration, a pressure sensor for detecting ground reaction force, and an electromyography (EMG) sensor for detecting muscle activation state; the IMU, pressure sensor, and EMG sensor are all electrically connected to the main controller.

[0011] Furthermore, the main controller is embedded with an algorithm system, which adopts a hierarchical structure, divided into a data layer, a feature layer, a decision layer, and a control layer from top to bottom; The data layer is responsible for calibrating, fusing, and extracting features from raw signals from inertial measurement, pressure, and electromyography sensors, providing preprocessed data for action intent recognition; The feature layer detects key gait events and constructs multi-dimensional feature vectors, transforming preprocessed sensor data into a quantitative description of the current motion state; Based on a finite state machine and preset rules, the decision layer maps the motion state provided by the feature layer into a specific push rod assist strategy, so that the intelligent action intention can be identified to assist decision-making. The control layer uses an impedance control model to convert the decision-making layer's assistance commands into specific motor control signals, thereby completing the assistance output.

[0012] Furthermore, the data layer is configured to integrate an inertial measurement unit, a pressure sensor, and an electromyography sensor. The inertial measurement unit collects triaxial acceleration and triaxial gyroscope data at a sampling rate of no less than 100Hz, and performs sensor fusion through complementary filters or Kalman filters to calculate the joint angle and angular velocity of the thigh and lower leg in the sagittal plane. Pressure sensors, distributed on the forefoot and heel of the shoe cover, with a sampling rate of no less than 100Hz, measure the ground reaction force and normalize the pressure value to a percentage of body weight, while calculating the trajectory of the center of pressure on the sole of the foot. An electromyography (EMG) sensor with a sampling rate of no less than 1000 Hz collects surface EMG signals from the quadriceps and hamstring muscles. The signals are then processed sequentially through bandpass filtering, full-wave rectification, and smoothing to obtain a linear envelope characterizing the level of muscle activation.

[0013] Furthermore, the feature layer is configured to perform the following operations: Based on data from pressure sensors and inertial measurement units, gait phases are segmented using a key event detection algorithm; Critical event detection includes: Heel contact event detection: When the value of the heel pressure sensor exceeds the first preset threshold and the angular velocity measured by the inertial measurement unit at the lower leg changes from a positive value to a negative value, it is determined as a heel contact event. Toe-off event detection: When the value of the palm pressure sensor drops below the second preset threshold and the acceleration measured by the inertial measurement unit at the thigh shows a specific peak value, it is determined as a toe-off event; Within each control cycle, a multidimensional feature vector is constructed to describe the current motion state. The multidimensional feature vector contains at least the knee joint angle calculated by the inertial measurement unit. With angular velocity heel pressure value Forefoot pressure value Foot pressure center trajectory And the electromyographic signal envelope.

[0014] Furthermore, the decision-making layer is configured to implement action intent recognition and decision-making assistance based on a finite state machine, wherein: The states of a finite state machine include at least: standing, walking swinging phase, walking support phase, squatting and standing up; The decision layer is configured to perform state transitions based on the feature vectors provided by the feature layer and according to preset state transition rules; The state transition rules include: when a toe-off event from the feature layer is detected, the state transitions from standing to the walking swing phase; when a heel-on event from the feature layer is detected, the state transitions from the walking swing phase to the walking support phase; when the knee joint angle is greater than a first angle threshold and the trunk forward tilt angle is greater than a first tilt angle threshold, the state transitions from standing to squatting; when the knee joint angular velocity is less than a first angular velocity threshold and the ground reaction force is greater than a first force threshold, the state transitions from squatting to standing. The decision layer is also configured to: query a predefined assist strategy mapping table based on the current state determined by the finite state machine, and generate assist decisions corresponding to the first electric actuator and the second electric actuator; the assist decisions include at least controlling the retraction of the second electric actuator during the middle of walking support, controlling the retraction of the first electric actuator during the early stage of walking swing, controlling the coordinated extension of the first and second electric actuators during the squatting process, and controlling the strong retraction of the first electric actuator during the standing process.

[0015] Furthermore, the control layer is configured to employ an impedance control model to convert the decision-making assistance commands generated by the decision layer into servo drive signals for the first and second electric actuators. The impedance control model is as follows: ; in, To achieve the desired joint output torque; and These are the actual joint angles and angular velocities calculated by the inertial measurement unit, respectively. The desired joint angle is set based on the finite state machine state of the decision layer. and These are the virtual stiffness coefficient and virtual damping coefficient, which determine the degree of compliance of the assist; The control layer is further configured to: based on the mounting positions of the first and second electric actuators on the leg support frame and the joint mechanics model, generate the desired joint output torque. Mapped to the required output force of the corresponding electric linear actuator ; output demand force As a given signal for the current loop, the servo motor inside the electric actuator is driven, and the closed-loop feedback formed by the position and force sensors built into the electric actuator is used to achieve precise and stable control of the output force.

[0016] Compared with existing technologies, the active-powered wearable assistive device provided by this invention, through its innovative overall structural design and intelligent control system, brings the following significant benefits: 1. Facilitates precise and efficient operation, significantly reducing workload: Through multimodal sensor fusion and intelligent decision-making based on finite state machines, the system can accurately and in real-time identify the user's intentions in walking, squatting, and standing. Employing an impedance control model for servo drive of the electric actuator achieves smooth, adaptive assistance rather than rigid mechanical actuation, ensuring a high degree of coordination between the assistance output and the user's own movements. In high-load activities such as bending over to lift heavy objects and climbing, it effectively reduces the load on key muscle groups in the legs and lower back by over 30%, directly alleviating work fatigue and reducing the risk of muscle and joint strain.

[0017] 2. Significantly Improved Ergonomics and Wearing Comfort: The overall structure employs an ergonomic three-point fixation system (chest-back-waist) and elastic straps, combined with a supportive back panel and lumbar support that conform to the body's curves. This ensures a close fit between the equipment and the body, with even force distribution, avoiding localized pressure points. The lower limb exoskeleton arms are driven by hinges and push rods, aligning with the knee joint's motion axis to ensure efficient and natural power transfer and reduce movement interference. The adjustable spring preload shock absorption mechanism effectively absorbs the load and impact vibrations generated when walking on uneven terrain, protecting the user's spine and further enhancing comfort during extended wear.

[0018] 3. Strong environmental adaptability and operational safety: Key structural components are encapsulated with insulating materials, and electrical isolation is considered in the overall design, meeting the safety regulations for live-line work in the power industry. The intelligent algorithm has real-time safety monitoring capabilities, promptly detecting abnormal states such as joint over-limit and push rod stall, cutting off power, and switching to passive safety mode to ensure user safety. Its modular and lightweight design allows it to adapt to various complex working environments such as mountainous areas and confined spaces.

[0019] 4. A good balance between range and practicality: By adopting a high-energy-density lithium battery pack and an efficient power management strategy, a range of over 4 hours of continuous operation is achieved while ensuring driving power, meeting the needs of most field shifts. Compared to existing bulky and short-range products, this equipment achieves a better balance between performance and practicality, making it more suitable for real-world engineering applications.

[0020] 5. Intelligent, controllable, and scalable: The core algorithm adopts a hierarchical architecture with a clear structure, facilitating subsequent function upgrades and algorithm optimization. The assist magnitude and response characteristics can be adjusted via software parameters, adapting to the personalized needs of users with different weights and work habits, demonstrating excellent versatility and scalability.

[0021] In summary, this invention effectively integrates mechanical structure, drive control, and intelligent algorithms, successfully solving the core pain points of existing technologies such as "bulky, short battery life, inaccurate recognition, and poor human-machine coupling." It provides an advanced wearable device that combines high-efficiency assistance, comfortable wear, safety and reliability, and long battery life, making it particularly suitable for high-intensity work scenarios such as power repair and emergency rescue. It has extremely high practical value and promising prospects for promotion. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0023] Figure 1 and Figure 2 This is an overall structural diagram of the active power wearable assistive device of the present invention; Figure 3 This is a structural diagram of the front chest controller, the main back support frame, and the lumbar support plate of the present invention; Figure 4 This is a structural diagram of the leg support frame, shoe cover, and power assist drive structure of the present invention; Figure 5 This is a structural diagram of the shoe cover of the present invention; Figure 6 This is a schematic diagram of the active power wearable assistive device of the present invention after it has been worn. Figure 7 This is a flowchart of the motion recognition and intelligent control algorithm of the present invention; In the diagram: 1. Front chest controller; 2. Back main support frame; 3. Lumbar support board; 4. Leg support frame; 5. Wearing shoe covers; 6. Power assist drive structure; 11. Chest safety harness; 12. Shoulder straps; 21. Support back panel; 22. Back frame; 23. Load-bearing platform; 24. Adjustable spring preload connector. 31. Waist safety harness; 41. Thigh exoskeleton arm; 42. Lower leg exoskeleton arm; 43. Thigh elastic band. 51. Adjustable shoe buckle; 61. First electric actuator; 62. Second electric actuator. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] Example 1 To address the problems of existing devices, such as bulkiness, short battery life, inaccurate recognition, and poor human-machine interaction, this embodiment provides an active-powered wearable assistive device to achieve the application goals of lightweight, long battery life, intelligent and precise assistance, safety and reliability, and comfortable wear.

[0026] like Figure 1 and Figure 2 As shown, the active power wearable assistive device adopts an exoskeleton-style wearable structure. Its core hardware, from top to bottom, mainly includes: a front chest controller 1, a back main support frame 2, a waist support plate 3, and a leg support frame 4 symmetrically arranged to assist the lower limbs, wearable shoe covers 5, and an assistive drive structure 6.

[0027] like Figure 1 , Figure 2 and Figure 3As shown in the figure, the front chest controller 1 is a flat box-shaped structure, fixed at the exact center of the wearer's chest. The front chest controller 1 highly integrates the core of the equipment inside, including a main controller equipped with an embedded system, a power supply system equipped with a high specific energy lithium battery pack, a signal processing module, and a wireless communication module. The front chest controller 1 is responsible for the operation, control, energy management, and data exchange of the entire system. The outer shell of the front chest controller 1 is integrally encapsulated with a high-insulation composite material to meet the safety requirements in the live working environment.

[0028] As Figure 1 , Figure 2 and Figure 3 shown in the figure, the main back support frame 2 forms the rigid backbone of the equipment, presenting a stable middle-character-shaped structure. The main back support frame 2 is mainly composed of a support backboard 21 and a pair of back frames 22. The shape of the support backboard 21 is an ergonomic curved panel that conforms to the physiological curvature of the human spine, directly contacting the user's back, and is used to evenly disperse the load and improve the wearing comfort. The back frame 22 is a rigid frame, fixedly connected to the support backboard 21, and symmetrically arranged on both sides of the support backboard 21. The front chest controller 1 is connected to the upper end of the back frame 22 through a pair of shoulder straps 12, forming a bearing structure that spans the shoulders; at the same time, the front chest controller 1 is connected to the support backboard 21 through a chest safety strap 11. Through the chest safety strap 11 and the safety buckle on it, the front chest controller 1 and the support backboard 21 are tightened and fixed in front of the chest, jointly forming a stable "chest-back" tension support system.

[0029] As Figure 1 , Figure 2 and Figure 3 shown in the figure, there are a pair of lumbar support plates 3, which are connected to both sides of the bottom of the support backboard 21 in a hinged manner, and their shapes conform to the curves on both sides of the human waist. A waist safety strap 31 is installed between the pair of lumbar support plates 3, and the waist safety strap 31 is located in the front of the human waist; the pair of lumbar support plates 3 and the waist safety strap 31 jointly act to firmly bind the equipment to the wearer's waist, which is the key node for the downward transmission of force. The lower end of the lumbar support plate 3 extends outward to the outer side of the human thigh and is hinged to the upper end of the leg support frame 4 of the lower limb.

[0030] The front chest controller 1, the main back support frame 2, and the lumbar support plate 3 constitute the wearing and fixing mechanism of this active power wearable assistive equipment, and the leg support frame 4, the wearable shoe cover 5, and the assistive driving structure 6 constitute the lower limb assistive mechanism of this active power wearable assistive equipment; the lower limb assistive mechanism is the core execution mechanism for realizing active assistance, and the lower limb assistive mechanism is designed symmetrically for the left and right legs, specifically as follows: As Figure 1 , Figure 2 and Figure 4As shown, the leg support frame 4 includes a thigh exoskeleton arm 41 and a calf exoskeleton arm 42. The thigh exoskeleton arm 41 is a rigid arm-shaped structure, with its upper end hinged to the lumbar support plate 3 and its lower end hinged to the calf exoskeleton arm 42 via a triangular connector. The thigh exoskeleton arm 41 is equipped with a thigh elastic strap 43 for securely binding the thigh exoskeleton arm 41 to the user's thigh. The calf exoskeleton arm 42 is also a rigid arm-shaped structure, linked to the thigh exoskeleton arm 41 via a triangular connector, which allows the knee joint to rotate flexibly in multiple degrees of freedom. The bottom of the calf exoskeleton arm 42 is hinged to a shoe cover 5.

[0031] like Figure 1 , Figure 2 and Figure 4 As shown, the power-assisted drive structure 6 includes a first electrically powered push rod 61 mounted on the thigh exoskeleton arm 41, and a second electrically powered push rod 62 mounted between the calf exoskeleton arm 42 and the shoe cover 5. The first electrically powered push rod 61 is hinged at both ends to the upper ends of the thigh exoskeleton arm 41 and the calf exoskeleton arm 42, respectively, spanning the knee joint. The first electrically powered push rod 61 primarily simulates the function of the quadriceps femoris muscle, providing the main extension torque during standing and climbing. One end of the second electrically powered push rod 62 is hinged to the lower middle section of the calf exoskeleton arm 42, and the other end is hinged to the heel of the shoe cover 5. The second electrically powered push rod 62 primarily simulates the function of the gastrocnemius muscle, providing auxiliary thrust during push-off from the ground while walking.

[0032] like Figure 4 and Figure 5 As shown, the foot fixation device includes a wearable shoe cover 5 and an adjustable shoe buckle 51. The wearable shoe cover 5 is a rigid base used to fix the user's shoe sole. The wearable shoe cover 5 is equipped with an adjustable shoe buckle 51, which can be adjusted and locked according to different shoe types and sizes to ensure effective force transmission between the foot and the exoskeleton structure and prevent relative displacement.

[0033] Furthermore, such as Figure 3 As shown, in addition to the wear fixation mechanism and lower limb assist mechanism, this powered wearable assistive device also features a load-bearing and shock-absorbing mechanism. This mechanism integrates carrying and cushioning functions to enhance work capacity and comfort. The load-bearing and shock-absorbing mechanism includes a load-bearing platform 23 and shock-absorbing components.

[0034] The load-bearing platform 23 is a flat plate. One side of the load-bearing platform 23 is hinged to the lower rear of the back frame 22, located between the user's back and the lumbar support 3. The load-bearing platform 23 is used to carry tools, spare batteries, or other heavy objects required for work. The shock-absorbing assembly is a set of spring-preload adjustable connectors 24 hinged between the load-bearing platform 23 and the lumbar support 3. The spring-preload adjustable connector 24 includes a spring, a preload adjustable connector, and a support column; the spring is the main shock-absorbing element, fitted onto the preload adjustable connector and support column, and its pre-compression can be adjusted by manually rotating the preload adjustable connector. The shock-bearing assembly effectively absorbs vertical impact forces generated by terrain undulations or heavy-duty walking, and filters vibrations, reducing the impact on the user's spine and joints.

[0035] In addition, this powered wearable assistive device is equipped with a sensor system to work in conjunction with the chest controller 1. The sensor system includes an inertial measurement unit (IMU), pressure sensors, and electromyography (EMG) sensors. The IMU is mounted in the middle of the thigh exoskeleton arm 41 and the lower leg exoskeleton arm 42, respectively, to measure limb acceleration and angular velocity in real time and calculate joint angles. Pressure sensors are embedded in the forefoot and heel of the wearable shoe cover 5 to detect the distribution of ground reaction force and gait phase. The EMG sensors are attached to the skin surface of the user's thigh to detect electrical signal activity of muscles, assisting in recognizing movement intentions. All sensors are electrically connected to the chest controller 1 via cables.

[0036] This powered wearable assistive device, when worn, is like... Figure 6 As shown, the various hardware components, through a sensor system, work in conjunction with the chest controller 1 and the power assist drive structure 6 to form an intelligent whole. The main controller within the chest controller 1 receives data from all sensors and uses a built-in intelligent algorithm to determine the wearer's intentions in real time, such as walking, squatting, and standing up. It then generates control commands to precisely drive the first electric push rod 61 and the second electric push rod 62 of the power assist drive structure 6 to extend and retract in tandem. This outputs timely and appropriate auxiliary torque at key stages of the movement, ultimately reducing the load on the user's muscle groups.

[0037] Example 2 Based on the hardware structure of the active-powered wearable assistive device in Example 1, this example elaborates on the motion recognition and intelligent control algorithms embedded in the main controller of the front chest controller. For example... Figure 7 As shown, this algorithm constitutes the intelligent hub of the equipment. Its core lies in achieving complete, closed-loop control from perception to execution, ensuring that the assist output and the human body's movement intentions are naturally, accurately, and smoothly coordinated. The overall architecture of the algorithm adopts a hierarchical structure, and its workflow constitutes an efficient perception-decision-execution closed loop, as follows: 1. Data Layer: The data layer performs multimodal signal acquisition and preprocessing. As the starting point for system perception, the data layer is responsible for purifying, calibrating, and initially fusing the raw signals from various sensors, providing a reliable data foundation for upper-level decision-making.

[0038] Signal Acquisition: Inertial Measurement Unit (IMU): An IMU deployed in the middle of the exoskeleton arm of the thigh and calf simultaneously acquires triaxial acceleration and triaxial gyroscope data at a sampling rate of at least 100 Hz. Pressure Sensors: Multiple pressure sensors distributed in the forefoot and heel of the worn shoe cover measure ground reaction force (GRF) at a sampling rate of at least 100 Hz. Electromyography (EMG) Sensors: EMG sensors attached to the quadriceps and hamstring muscles acquire surface electromyography signals at a sampling rate of at least 1000 Hz.

[0039] Preprocessing includes: IMU signal processing, pressure signal processing, and EMG signal processing, as detailed below: For IMU signal processing: First, perform data calibration: , ; in, , Let be the calibrated acceleration and angular velocity vectors at time t; , This is the original reading vector; , It is a pre-calibrated zero bias vector.

[0040] Then sensor fusion is performed: ; ; ; in, The angle calculated by the accelerometer; It is a bivariate arctangent function; , , These are the calibrated acceleration values ​​of the Y-axis, X-axis, and Z-axis at time t. To calculate the projection length of the acceleration vector onto the XZ plane; This is a constant used to convert radians to degrees. The integral angle of the gyroscope; The final fusion angle output by the complementary filter at the previous time (t-1); Let t be the calibrated Y-axis angular velocity; The sampling period. This refers to the angle of the fused knee joint; These are the filter coefficients; The weights of the accelerometers.

[0041] For pressure signal processing: Normalization of ground reaction force: ; in, Normalized ground reaction force; , The pressure values ​​are for the heel and forefoot. The user's weight.

[0042] Furthermore, based on the pressure distribution between the forefoot and heel, the trajectory of the plantar pressure center is calculated in real time. : ; in, Location of the pressure center; and These represent the lever arm distances from the forefoot and heel sensors to the ankle hinge point, respectively. Foot pressure center trajectory. Used to determine the rolling state of the foot.

[0043] For EMG signal processing: Full-wave rectification: ; in, The rectified signal is the result of full-wave rectification of the original bandpass filtered electromyographic signal. The signal is the bandpass filtered signal, which is the output result of the original electromyography signal after bandpass filtering.

[0044] Envelope extraction: ; in, Electromyographic envelope; This is the window length corresponding to a 200ms time window.

[0045] 2. Feature layer: The feature layer performs motion feature extraction and gait event detection. It extracts key events and state feature vectors from preprocessed data, with the core task being accurate segmentation of the gait phase.

[0046] The feature layer sets up a key event detection algorithm, including: heel touches the ground When the heel pressure sensor reading exceeds the first preset threshold and the angular velocity measured by the calf IMU changes from a positive value to a negative value, it is determined as a heel contact event. Heel touching the ground indicates: ; in, The threshold for judging heel strike events is usually set at 10%-20% of body weight; The value is the angular velocity of the knee joint in the lower leg.

[0047] toes off the ground When the current palm pressure sensor value drops below the second preset threshold and the thigh IMU acceleration shows a specific peak value representing forward swing, it is determined as a toe-off event.

[0048] The toes leaving the ground is represented as: ; in, The threshold for determining a toe-off event is typically set at 5% of body weight. The acceleration in the forward direction at the thigh; This is the peak acceleration threshold.

[0049] Feature vector construction: in each control cycle Construct feature vectors : ; in, for The feature vector at time step; Indicates vector transpose; Let be the knee joint angle at time t; Let be the angular velocity of the knee joint at time t; The envelope of the quadriceps femoris electromyography signal at time t represents the real-time activation level of the quadriceps femoris on the front of the thigh and is a direct biosignal for judging the intention of actions such as standing up and pushing off. This is the hamstring electromyographic signal envelope, representing the real-time activation level of the hamstrings on the posterior thigh, and can be used to assist in assessing knee flexion stability during the swing phase or stability during the stance phase. (Eigenvector) It is a snapshot of the current motion state and serves as the direct basis for decision-makers to make state judgments.

[0050] 3. Decision-making level: The decision-making layer identifies action intent and assists in decision-making. The decision-making layer is the intelligent core of the algorithm, and it implements action recognition and decision-making through a finite state machine (FSM).

[0051] Finite state machine design and state transition: System state set ;in, The system's set of states; Standing position describes the wearer in a static or dynamic upright posture, with both feet or one foot stably supporting the body, without any obvious walking or squatting movements; The gait swing phase describes the stage in the gait cycle where one leg is swinging forward in the air. The support phase of walking describes the stage in the walking cycle where one leg is in contact with the ground and supporting the body weight. The description indicates a squatting position, where the wearer is bending their knees and lowering their center of gravity. The description is of a standing position, indicating that the wearer is extending their knees and raising their center of gravity.

[0052] State transition is determined by the feature vector The triggering rules are as follows: The swinging phase from standing to walking: when the system is in a standing state If a toe-off event is detected... =1, then the state transitions to the walking swaying phase. .

[0053] From the swaying phase to the support phase: When the system is in the swaying phase If a heel-to-ground event is detected... =1, then the state transitions to the walking support phase. .

[0054] From standing to squatting: When the system is in a standing position At that time, if the knee joint angle satisfies >45° and the torso lean angle meets the requirements. If the angle is greater than 30°, the state transitions to squatting. .

[0055] From squatting to standing: When the system is in a squatting state If the angular velocity of the knee joint satisfies This indicates that the extension and the ground reaction force satisfy... If the percentage is greater than 50%, the state will transition to "Stand Up". .

[0056] From the walking support phase to standing: When the system is in the walking support phase At that time, if the center of pressure on the sole of the foot meets the requirements >0.6 If the center of pressure exceeds 60% of the foot length, the state shifts to standing. Entering the final stage of standing.

[0057] From standing up to standing: When the system is in the standing state At that time, if the knee joint angle satisfies <10° and knee joint angular velocity approximately zero Then the state transitions to standing. Assisted Decision Mapping: Once the state is determined, the system queries a predefined assist strategy mapping table to generate the corresponding lever control command. Walking support phase assistance strategy: When the system is in In this state, control the second electric actuator to apply force. The contraction mimics the exertion of the gastrocnemius muscle, assisting in pushing off the ground.

[0058] Assist strategy during the walking sway phase: When the system is in In this state, control the first electric actuator to apply force. Contraction assists in the forward swing of the thigh, conserving energy for leg lifting.

[0059] Squatting process assistance strategy: When the system is in In this state, control the first and second electric actuators to apply force. Coordinated stretching provides controlled resistance, assists in slow squatting, and protects the knee joint.

[0060] Strategy to assist in the standing process: When the system is in In this state, control the first electric actuator to apply maximum force. Contraction mimics the function of the quadriceps, providing the main power for standing up and significantly reducing the load on the thighs.

[0061] 4. Control Layer: The control layer provides servo drive and compliant assist output. The control layer translates the abstract instructions from the decision layer into compliant physical assistance that the motor can execute.

[0062] The control layer is configured to use an impedance control model to convert the decision-making assistance commands generated by the decision layer into servo drive signals for the first and second electric actuators. The impedance control model is as follows: ; in, To achieve the desired joint output torque; and These are the actual joint angles and angular velocities calculated by the inertial measurement unit, respectively. The desired joint angle is set based on the finite state machine state of the decision layer. and These are the virtual stiffness coefficient and virtual damping coefficient, which determine the degree of compliance of the assist.

[0063] The torque mapping calculation formula for the impedance control model is as follows: ; in, To provide the required output force for the push rod; The lever arm from the pushrod mounting point to the joint; The instantaneous angle between the push rod and the skeletal arm.

[0064] The servo drive calculation formula for the impedance control model is as follows: ; in, Set the current loop value for the servo driver; This is the constant force of the push rod. The position and force sensors built into the push rod form a closed-loop feedback to ensure the accuracy and stability of the output force.

[0065] 5. Algorithm Implementation and Security Assurance: Real-time performance guarantee: The entire algorithm runs on a high-performance embedded platform, such as ARM Cortex-M7, ensuring that the entire process control cycle from signal acquisition to torque output is ≤10ms, meeting the requirements for real-time response.

[0066] Adaptive adjustment: The system can record the user's average EMG level under different activities and fine-tune the boost gain accordingly. This allows for personalized support intensity.

[0067] Safety Monitoring: The system has a dedicated safety monitoring thread that continuously monitors sensor data and system status. Once an exceedance is detected, [the system will take immediate action]. >120° or stall If the sensor signal is lost for more than 200ms, the motor power will be immediately cut off, switching the equipment to passive following mode to maximize user safety. The actual output force of the push rod at time t; This is the maximum allowable force deviation tolerance.

[0068] The motion recognition and intelligent control algorithm embedded in the main controller of the front chest controller provided in this embodiment starts from multimodal signal acquisition and preprocessing, through feature extraction and gait event detection, through intelligent decision-making based on finite state machines, and finally achieves compliant assist output through impedance control model, forming a complete, closed-loop, and detailed and feasible technical path.

[0069] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An active powered wearable assistive equipment, characterized by: The active power wearable assisting equipment comprises a front chest controller, a back main support frame, a waist support plate, a leg support frame, a wearable shoe cover and an assisting driving structure; The front chest controller is integrated with a main controller, a power supply and a signal processing module, and is connected with the back main support frame through a group of shoulder straps and a group of chest safety straps; the back main support frame can be fitted with the back curve of the human body, and a group of waist support plates are fixedly installed at the bottom of the back main support frame; the group of waist support plates can be fitted with the waist curve of the human body, and are respectively connected with two groups of leg support frames in the downward direction, and a waist safety strap is installed between the group of waist support plates and at the front side of the waist of the human body; the bottom of the two groups of leg support frames is provided with the wearable shoe cover; The assisting driving structure comprises a first electric push rod installed on the thigh exoskeleton arm and a second electric push rod installed between the lower leg exoskeleton arm and the wearable shoe cover, and the first electric push rod and the second electric push rod are electrically connected with the front chest controller; the front chest controller can realize real-time sensing of the motion state and action intention of the wearer through the sensor system integrated in the equipment, and can realize cooperative control of the output force, speed and stroke of the first electric push rod and the second electric push rod, so as to provide timely and appropriate auxiliary torque in the key stages of walking, bending and standing, and to complete the assisting.

2. The active powered wearable assistive equipment of claim 1, wherein: The front chest controller adopts an insulation packaging structure to insulate and package the main controller, the power supply and the signal processing module; The back main support frame has a middle character-shaped structure, which comprises a support back plate and a group of back frames fixedly connected with the support back plate; the support back plate is connected with the front chest controller through a group of chest safety straps, and buckles are arranged on the chest safety straps; the group of back frames is connected with the front chest controller through a group of shoulder straps.

3. The active powered wearable assistive equipment of claim 2, wherein: The waist support plate is hinged to the support back plate at the back side of the waist of the human body, and the lower end extends to the outside of the two sides of the thighs of the human body and is hinged to the leg support frame; A load platform is further installed between the back main support frame and the group of waist support plates; the load platform is a flat plate body, which is located at the back side of the human body and is hinged to the support back plate and the group of back frames, and a spring preloading adjustable connecting piece is hingedly installed between the load platform and the group of waist support plates.

4. The active power wearable assistive equipment of claim 1, wherein: The leg support frame comprises a thigh exoskeleton arm and a lower leg exoskeleton arm; one end of the thigh exoskeleton arm is hinged to the waist support plate, the other end is hinged to one end of the lower leg exoskeleton arm, and the other end of the lower leg exoskeleton arm is hinged to the wearable shoe cover; One end of the first electric push rod of the assisting driving structure is hinged to the upper end of the thigh exoskeleton arm, and the other end is hinged to the upper end of the lower leg exoskeleton arm; one end of the second electric push rod is hinged to the middle and lower segment of the lower leg exoskeleton arm, and the other end is hinged to the rear root of the wearable shoe cover; the thigh exoskeleton arm is provided with a thigh elastic band that is fixed to the thigh of the human body; An adjustable shoe buckle is arranged on the wearable shoe cover to facilitate wearing.

5. The active power wearable assistive equipment of claim 1, wherein: The sensor system comprises an inertial measurement unit (IMU) for collecting leg angular velocity and acceleration, a pressure sensor for detecting ground reaction force, and an electromyography (EMG) sensor for detecting muscle activation state; the inertial measurement unit (IMU), the pressure sensor and the electromyography (EMG) sensor are electrically connected with the main controller.

6. The active power wearable assistive equipment of claim 1, wherein: The main controller is embedded with an algorithm system, which adopts a hierarchical structure and is divided into a data layer, a feature layer, a decision layer and a control layer from top to bottom; The data layer is responsible for calibrating, fusing and extracting features from the raw signals from the inertial measurement, pressure and electromyographic sensors, and provides preprocessed data for action intention recognition; The feature layer converts the preprocessed sensor data into a quantitative description of the current motion state by detecting key gait events and constructing a multi-dimensional feature vector; The decision layer maps the motion state provided by the feature layer into a specific putter assistance strategy based on a finite state machine and preset rules, so that the intelligent action intention is recognized and assistance decisions are made; The control layer converts the assistance instructions from the decision layer into specific motor control signals through an impedance control model to complete the assistance output.

7. The active power wearable assistive equipment of claim 6, wherein: The data layer is configured to integrate an inertial measurement unit, a pressure sensor and an electromyographic sensor; The inertial measurement unit collects three-axis acceleration and three-axis gyroscope data at a sampling rate of not less than 100 Hz, and performs sensor fusion through a complementary filter or a Kalman filter to calculate the joint angle and angular velocity of the thigh and calf links in the sagittal plane; The pressure sensor is distributed on the forefoot and heel of the wearable shoe cover, and has a sampling rate of not less than 100 Hz, measures the ground reaction force, and normalizes the pressure value to a percentage of body weight, while calculating the trajectory of the center of pressure on the sole; The electromyographic sensor has a sampling rate of not less than 1000 Hz, collects the surface electromyographic signals of the quadriceps femoris and hamstrings, and sequentially passes through band-pass filtering, full-wave rectification and smoothing processing to obtain a linear envelope line representing the muscle activation level.

8. The active powered wearable assistive equipment of claim 6, wherein: The feature layer is configured to perform the following operations: based on the data from the pressure sensor and the inertial measurement unit, the gait phase is segmented through a key event detection algorithm; Key event detection includes: heel strike event detection: when the heel pressure sensor value exceeds a first preset threshold, and the angular velocity measured by the inertial measurement unit at the calf changes from positive to negative, it is determined as a heel strike event; toe-off event detection: when the forefoot pressure sensor value drops below a second preset threshold, and the acceleration measured by the inertial measurement unit at the thigh appears a specific peak value, it is determined as a toe-off event; In each control cycle, a multi-dimensional feature vector is constructed to describe the current motion state The multi-dimensional feature vector contains at least the knee angle calculated by the inertial measurement unit The angular velocity The heel pressure value The forefoot pressure value The center of pressure trajectory of the foot bottom And the envelope of the myoelectric signal.

9. The active power wearable assistive equipment of claim 6, wherein: The decision layer is configured to realize action intention recognition and assistance decision based on a finite state machine, wherein: The states of the finite state machine include at least: standing, walking swing period, walking support period, squatting and standing up; The decision layer is configured to perform state transition according to the state transition rules based on the feature vector provided by the feature layer; The state transition rules include: when the toe-off event from the feature layer is detected, the state is transferred from standing to walking swing period; when the heel strike event from the feature layer is detected, the state is transferred from walking swing period to walking support period; when the knee joint angle is greater than a first angle threshold and the trunk forward inclination angle is greater than a first inclination threshold, the state is transferred from standing to squatting; when the knee joint angular velocity is less than a first angular velocity threshold and the ground reaction force is greater than a first force threshold, the state is transferred from squatting to standing up; The decision layer is further configured to: according to the current state determined by the finite state machine, query a predefined assist strategy mapping table to generate an assist decision corresponding to the first electric push rod and the second electric push rod; the assist decision at least includes controlling the second electric push rod to contract in the walking support middle stage, controlling the first electric push rod to contract in the walking swing early stage, controlling the first and second electric push rods to coordinate expansion in the squatting process, and controlling the first electric push rod to strongly contract in the standing process.

10. The active power wearable assistive equipment of claim 6, wherein: The control layer is configured to use an impedance control model to convert the assist decision instruction generated by the decision layer into servo driving signals for the first electric push rod and the second electric push rod. The impedance control model is: ； wherein, is the desired joint output torque; and are the actual joint angle and angular velocity, respectively, as resolved by an inertial measurement unit; is the desired joint angle set according to the finite state machine state of the decision layer; and are the virtual stiffness and virtual damping coefficients, respectively, that determine the degree of assist compliance; The control layer is further configured to: according to the installation positions of the first electric push rod and the second electric push rod on the leg support frame and the joint mechanics model, map the expected joint output torque to a demand output force of the corresponding electric push rod ; take the demand output force as a given signal of a current loop, drive a servo motor in the electric push rod, and realize accurate and stable control of the output force by using a closed-loop feedback formed by a position and force sensor built in the electric push rod.

Images