Hip-assisted exoskeleton control method based on force-position information fusion

Abstract

The application relates to a hip power-assisted exoskeleton control method based on force-position information fusion, which accurately identifies different motion states through multi-modal perception and fusion, and realizes intelligent switching of multiple power-assisted modes, comprising the following steps: identifying different states through multiple source data such as the pitch angle and angular velocity of the body back IMU and the hip motor encoder angle value. Corresponding control is carried out in different motion states: a zero force drag algorithm is adopted in the walking state to reduce resistance and torque compensation is carried out based on the hip joint speed to enhance followability; a power-assisted algorithm combining PD control based on position error and support force compensation is adopted in the standing-up state; support force is provided in the squatting and standing stop states. The control part comprises a data processing module, a state recognition module, a control strategy module and a motor control module. The application can accurately identify the motion state, effectively improve the power-assisted efficiency, reduce the physical energy consumption of the wearer and adapt to various task scenarios.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of exoskeleton control, and particularly relates to a hip-assisted exoskeleton control method based on force-position information fusion, which can accurately identify human motion patterns and provide efficient and accurate assistance control for hip-assisted exoskeletons, thereby improving the cooperative motion performance of the exoskeleton and the human body. BACKGROUND

[0002] With the continuous progress of science and technology, exoskeleton robots have shown an increasingly broad application prospect in many fields such as logistics, industrial manufacturing, old-age care, and construction. It has important significance in assisting human to carry heavy objects and strengthening human motion ability. However, the current assisted exoskeleton technology still faces many severe challenges.

[0003] Early exoskeleton control is mainly pre-programmed control. For example, the "ReWalk" developed by the Israeli "exoskeleton" assistance device ReWalk Elg Medical Technology Company helps maintain body balance with a pair of crutches, is composed of an electric leg support, a body sensor, and a backpack, and the backpack has a computer control box and a rechargeable battery. Users can use a remote waistband to select a certain setting, such as standing, sitting, walking, and climbing, and then lean forward to activate the body sensor to put the mechanical leg into motion. It is mainly used to help paraplegics recover walking ability. The power transmission adopts the method of motor-reducer-exoskeleton mechanism, and the motion mode is mainly that the device drives the human body to move, and the size of the assistance of the device is set by the control system and cannot be changed at any time according to the action intention of the human body.

[0004] The Berkeley Lower Extremity Exoskeleton (BLEEX) is a device designed by the Defense Advanced Research Projects Agency (DARPA) to try to link a robotic support to a person's legs to reduce the weight load, thus enabling infantry to travel longer distances with heavier loads. The device consists of a fuel supply and engine system, control and sensing system, hydraulic transmission system and exoskeleton mechanism. The person using the device connects his legs to the mechanical exoskeleton via a transmission belt and carries a large backpack containing the engine and control system. The backpack also has space to carry a payload. The power transmission process is: engine - hydraulic system - exoskeleton mechanism. The device balances its own weight (50 kg) so that the person wearing it feels no load and the control system ensures that its center of gravity is always on the user's feet. The backpack can also carry a 32 kg load, while the user feels as if he is carrying only 2 kg. The device can help soldiers, but also medical personnel to evacuate the wounded from dangerous areas or firefighters to climb more floors with heavy equipment.

[0005] Wearable robotic suits will be able to help troops to increase their fighting and endurance capabilities. The Raytheon Sarcos model developed by Raytheon has a claw-like hand. When American soldiers wear this equipment, their strength and endurance will be 20 times greater than normal. With the HULC model of Lockheed Martin, the wearer can easily carry 200 pounds (about 91 kg) and can perform more demanding tasks with less energy consumption. Currently, the US Army Soldier Systems Center is conducting military tests on robotic exoskeletons.

[0006] Another military cooperation project in the United States, Raytheon Sarcos XOS, which represents the latest level of exoskeleton robots, is the pride of Dr. Steve Jacobsen's mobile exoskeleton "XOS". The exoskeleton "XOS" was created to create a super soldier, and a military research budget of $10 million was provided by the U.S. Department of Defense Advanced Research Projects Agency (DARPA), and after 7 years of secret development, it represents the most advanced technology in the field of mechanical exoskeletons. Its control idea is the same as BLEEX, the control system determines how much assistance and speed to give the human body by detecting the system and the microcomputer system to determine the next action of the human body, and also transmits force to the exoskeleton mechanism through the hydraulic system, but it is a fully armed exoskeleton, while BLEEX is a lower limb exoskeleton robot. "XOS" moves much more agilely and powerfully than previous exoskeleton devices. Using sensors attached to the body, it can react to body movements without delay and output powerful force. When wearing "XOS", it can lift 90.7 kg of weight while feeling only 9 kg, and can lift 50-500 times continuously. But currently "XOS" has a major flaw, the battery it carries can only be used for 40 minutes.

[0007] According to the above information, the control technology of the military exoskeleton is advanced, but because it is a military secret technology, its control method is not disclosed.

[0008] The Chinese invention patent with publication number CN112949676B, which was granted on July 8, 2022, discloses a self-adaptive motion mode recognition method for a flexible lower limb assistance exoskeleton robot, comprising the following steps: 1) obtaining a large amount of swing state sensor data corresponding to various motion modes as human motion information; 2) extracting motion features according to the motion information, and selecting motion feature information corresponding to the minimum value time of the calf IMU; 3) inputting the normalized motion feature information into the BP neural network for training, and the output result is the recognition probability of each motion mode, and the larger the recognition probability is, the final motion mode recognition result is; 4) obtaining the feature vector corresponding to the minimum value time of the calf IMU; 5) using the trained neural network model to recognize the motion mode; 6) using the finite state transition algorithm to correct the recognition result of the BP neural network. The present application distinguishes the curve information position of the motion mode, and recognizes the motion mode according to the feature information of the position, thereby improving the motion mode.

[0009] 2024.03.08 published, the disclosure number CN117653505A Chinese invention patent discloses a kind of hip joint power-assisted exoskeleton robot, hip joint power-assisted method, equipment and medium, comprising: waistband, the left and right ends of waistband are provided with waist support piece;Two hip joint drive modules, two hip joint drive modules are all included rotary drive assembly and hip joint connecting piece, the left and right ends of rotary drive assembly are respectively connected with waist support piece and hip joint connecting piece;Two leg wear components, two leg wear components are all included upper leg bar and lower leg bar, the left and right ends of upper leg bar are respectively connected with hip joint connecting piece and lower leg bar, and lower leg bar is provided with leg belt;Control box, set in the front end of waistband, the left and right ends of control box are respectively connected with waist support piece by chain, controller is set in control box, and rotary drive assembly is electrically connected with controller;The embodiment of the application can be suitable for different body shape of user, and effectively provide power assistance to the hip joint of user. SUMMARY

[0010] In view of the above problems existing in the prior art, a hip power-assisted exoskeleton control method based on force-position information fusion is proposed. The human body motion pattern can be accurately identified, and efficient and accurate power assistance control is provided for the hip power-assisted exoskeleton, thereby significantly improving the cooperative motion performance of the hip power-assisted exoskeleton with the human body, effectively overcoming the shortcomings of the prior art, and greatly improving the performance and practicality of the power-assisted exoskeleton. The specific content is as follows:

[0011] The hip power-assisted exoskeleton control method based on force-position information fusion of the present application comprises the following steps:

[0012] S1. Multi-modal information acquisition and preprocessing

[0013] The body pitch angle and pitch angular velocity information are collected by the inertial sensor installed on the back, the hip joint angle information is collected by the absolute value encoder, and the hip joint torque information is collected by the torque sensor. The collected inertial sensor data is filtered to remove noise interference, and the absolute value encoder data is calibrated to ensure the accuracy of angle measurement. The torque sensor data is zero drift corrected,

[0014] S2. Motion state recognition and judgment

[0015] The body pitch angle and angular velocity collected by the inertial sensor are combined with the angle value of the absolute value encoder, and the state machine switching judgment is made according to the preset threshold condition. The state includes standing / walking state, squatting state, squatting stopping state, standing up state, standing up stopping state.

[0016] The body pitch angle threshold θ stand , the body pitch angular velocity threshold v move , the hip position difference threshold θ e, hip angular velocity product threshold v e , the motion state includes standing / walking state, squatting state, squatting stop state, standing up state, standing up stop state, and the specific determination method is as follows:

[0017] When the upper body is bent more than θ stand , and the bending angular velocity reaches v move , and the encoder position difference is less than θ e , and the left and right angular velocity product is greater than v e , switch from the standing state / walking state of both feet to the squatting state,

[0018] When the upper body stops bending during squatting, i.e. the upper body angular velocity is less than v move , and the left and right angular velocity product is less than v e , enter the squatting stop state,

[0019] When the squatting stop state or the standing up stop state, if the body is standing up, i.e. the upper body angular velocity is less than v move , or the angular velocity product is greater than v e , enter the standing up state,

[0020] When the upper body stops lifting during standing up, i.e. the upper body angular velocity is greater than v move , or the angular velocity product is less than v e , enter the standing up stop state,

[0021] When the body continues to stand up in the standing up stop state, it meets the standing up state condition, and the standing up state is maintained; when the body angle is less than θ stand , and the upper body angular velocity is less than v move , switch from the standing up state to the standing state / walking state of both feet.

[0022] As an optimization, the pitch angle and angular velocity data of the user's back inertial sensor, and the angle value data of the absolute value encoder are collected in real time to identify and determine the motion state and adopt different assistance strategies in different motion states. The control algorithm and execution in different states are as follows:

[0023] ① If in the standing / walking state, execute the walking assistance algorithm, including zero force drag algorithm to calculate the torque to overcome gravity and friction, and torque compensation algorithm based on hip joint speed to calculate compensation torque, and add the two to get the final control torque,

[0024] ② If in the standing up state, calculate the assistance torque according to the PD control law of the standing up assistance, and calculate the final control torque combined with the support force compensation,

[0025] ③ If in the squatting stop or standing up stop state, only calculate the control torque of the support force compensation part.

[0026] As optimization, during the whole control process, an error processing mechanism for sensor data jump is also included, specifically:

[0027] The data processing module processes the sensor values, including taking the left hip joint encoder value to the opposite to unify the left and right directions, differentiating the encoder values of adjacent sampling points and calculating the hip angular velocity, and zeroing the IMU. At the same time, if the current value and the value of the previous two sampling points change too much, the variable error is set to 1, indicating data jump,

[0028] The state recognition module determines the motion state according to the sensor data. When error is 1, the state is cut to initial 0, and the assist motor stops moving to ensure the safety of the wearer,

[0029] As optimization, the motor control module operation includes motor initialization and sending current value, specifically: the assist motor initialization is completed within 3 seconds after the exoskeleton robot is powered on, including starting the left and right hip assist motors, and sending instructions to set the speed limit of the motor,

[0030] After 3 seconds of power-on, the motor control module only performs the sending current value operation, converts the control torque value calculated by the control strategy module into current, and sends it to the driver, so that the assist motor runs under the current loop according to the control torque, realizes the assist action of the exoskeleton on the human hip, assists the human body to complete the action of bending and standing up, reduces the burden on the human body, and reduces the risk of muscle fatigue or injury. At the same time, the sensor data is continuously monitored, and the control strategy and torque output are adjusted in real time according to the state change.

[0031] As optimization, it includes two control modes:

[0032] Follow-up control mode:

[0033] According to the torque change caused by the instantaneous action of the user, the torque change sensed by the left torque sensor or / and the right torque sensor is determined, the user's left or / and right hip action trend is determined, and the left or / and right hip joint assist mechanism is controlled to follow its action;

[0034] II. Left and right synchronous control mode:

[0035] According to the angle, angular velocity, angular acceleration and assist torque sensed by the real-time measurement of the load inertia sensor, the user's trunk posture is determined, and then combined with the left and right hip joint assist mechanism angle and angle change measured by the absolute value encoder of the left and right hip joint assist mechanism, the current action type of the user is inferred, and the left and right hip joint assist mechanism is controlled to make corresponding action to complete the assist. Detailed judgment is as follows:

[0036] If the user's upper body is basically kept vertical, the left and right hip joint assistance mechanisms are alternately bent / extended, and it is determined that the user is in a walking state. The left and right hip joint assistance mechanisms have the same amplitude and speed but opposite directions.

[0037] If the user's upper body is forward inclined, the left and right hip joint assistance mechanisms are simultaneously extended, and it is determined that the user is in a standing-up state. The left and right hip joint assistance mechanisms have the same amplitude, speed and direction.

[0038] If the user's upper body is left inclined, the left hip joint assistance mechanism is bent, and the right hip joint assistance mechanism is extended, and it is determined that the user turns left. The left and right hip joint assistance mechanisms have opposite directions, and the left hip joint assistance mechanism has a smaller amplitude and speed than the right hip joint assistance mechanism.

[0039] If the user's upper body is right inclined, the right hip joint assistance mechanism is bent, and the left hip joint assistance mechanism is extended, and it is determined that the user turns right. The left and right hip joint assistance mechanisms have opposite directions, and the right hip joint assistance mechanism has a smaller amplitude and speed than the left hip joint assistance mechanism.

[0040] If the user's upper body is forward inclined, the hip angle value increases, and the upper body bending angular velocity reaches a set threshold v move , and the difference between the angle values measured by the absolute encoders at the two hip motors is less than a threshold θ e , and the product of the left and right angular velocities is greater than a threshold v e , it is determined that the user starts to squat.

[0041] When in the squatting stop state or the standing-up stop state, if the user's upper body starts to backward incline (the upper body motion angular velocity is less than v move or the product of the angular velocities is greater than v e ), it is determined that the user enters the standing-up state. During the standing-up process, the upper body continuously backward inclines, and the hip angle gradually decreases. The system calculates the assistance torque based on the position error PD control and combines the support force compensation to provide upward assistance for the user. The assistance torque is calculated according to the real-time hip angle and angular velocity data, provides a larger assistance to overcome gravity in the early stage of standing up, and gradually decreases with the body gradually upright. When the upper body stops lifting (the upper body angular velocity is greater than v move or the product of the angular velocities is less than v e ), the system continues to provide support force to ensure that the body is stable in the upright position, and continuously monitors the sensor data to determine whether further action adjustment is needed.

[0042] When the left-right synchronous control mode condition is met, the left-right synchronous control mode is adopted, otherwise the follow-up control mode is adopted.

[0043] 1. Hardware system composition

[0044] The exoskeleton robot includes a back structure, a waist structure, an upper limb wearing structure, a hip joint movement module and a leg binding mechanism. The back structure accommodates electrical elements such as batteries and controllers, provides support and assistance; the waist structure is adjustable, suitable for different body types; the upper limb wearing structure combines inner flexible material with outer rigid bearing structure, improving wearing comfort; the hip joint movement module is composed of a power-assisted motor, an encoder, a joint connecting disc and a thigh rod, realizing the movement control of the hip joint; the leg binding mechanism makes the exoskeleton leg rod closely fit the human leg.

[0045] The control system includes a motion control board (CSPACE-V4), a torque sensor, an external absolute value encoder, an IMU inertial sensor, a lithium battery and a DC-DC power converter.

[0046] 2. Multi-modal information acquisition and processing

[0047] ① IMU inertial sensor: installed in the shaft position of the back frame device, used for detecting the spatial posture of the exoskeleton back frame and determining the current exoskeleton gait mode. The sensor adopts MEMS technology, integrating three-axis gyroscope, three-axis accelerometer and motion sensor. It realizes accurate measurement and analysis through self-developed attitude dynamics core algorithm engine and high dynamic Kalman filtering fusion algorithm. It can directly output inclination angle, with high measurement accuracy and good stability, and can provide up to 200Hz update rate of data in real time, meeting various high-precision application requirements, widely used in industrial inclination measurement and other fields, with strong anti-electromagnetic interference ability, and can adapt to industrial harsh environment.

[0048] ② Torque sensor: mainly detects the joint torque of the exoskeleton hip joint and knee joint, rated voltage DC-7.5~15V, rated current ≤150mA, transmits torque data in real time through RS485 communication, interface protocol is RS-485 bus, baud rate is 115200, elastomer material is stainless steel, measurement range is ±300N.m, comprehensive measurement accuracy is ±0.5%F.S, creep is 0.1%F.S30min, working environment temperature is 0℃~+45℃, storage and transportation environment temperature is -20℃~+85℃.

[0049] ③Encoder: Each assistive motor is equipped with an absolute encoder in addition to the internal incremental encoder. The rated voltage is DC-7.5-15V, the rated current is ≤150mA, the communication interface is RS485, the interface protocol is RS-485 bus, the default baud rate is 115200, the measurement angle range is 0-360°, the resolution is 14Bit (0.0219° / LSB), the maximum rotation speed is 1200rpm, the power-up time is 100ms, the input shaft tolerance is <0.3mm, the magnetic flux is ±30mT...±70mT, and the typical magnet-to-encoder chip distance is 0.5mm-2.5mm.

[0050] 3. Control algorithm implementation

[0051] State machine design and switching logic: The state machine starts from the standing / walking state, including the squatting state, squatting stop state, standing up state, and standing up stop state. The switching conditions are based on the body pitch angle, body pitch angular velocity measured by the body back IMU, and the angle value and angular velocity product of the absolute encoder at the two hip motors, as well as other multi-modal information. For example, the condition for the standing / walking state → squatting state is that the upper body bending is greater than θ stand , the bending angular velocity reaches v move , the encoder position difference is less than θ e , and the left-right angular velocity product is greater than v e .

[0052] ①Zero force drag algorithm based on torque compensation (applicable to standing / walking state)

[0053] The robot driver works in current / torque mode, and the control torque is , where the gravity term is simplified as a function related to the trigonometric function of the joint angle θ . sin(θ) τ G =Ksin(θ) The joint friction torque is obtained, and thus the control torque of the knee and hip joints during swing is τ knee and τ hip The controller parameters K i (i=1,2…7) are determined by mass, center of mass estimation, and real machine test tuning to offset the gravity and achieve a compliant human-robot interaction force.

[0054] ②Torque compensation algorithm based on hip joint speed (applicable to standing / walking state)

[0055] ​Two sampling intervals are set, large sampling interval (more than 10 sampling points) is used to judge the acceleration or deceleration process of the hip joint movement, and determine the direction of the control torque; small sampling interval (less than 5 sampling points) is used to calculate the torque compensation value, and compensate the hip torque according to the current angle and speed, and enhance the following of the exoskeleton.

[0056] ③PD control based on position error (applicable to standing up state)

[0057] PD control based on position error is adopted, and the control law is

[0058]

[0059] wherein e(t)=θ d - θ is the error, wherein θ d is the desired angle, θ is the real-time measured motor angle value. K P is the proportional coefficient, K D is the differential coefficient. When standing up assistance is performed, the desired angle can be set to 0, and the appropriate assistance torque can be obtained by adjusting K P and K D .

[0060] ④Support force compensation (applicable to squatting stop state, standing up stop state and standing up state)

[0061] The upper part of the exoskeleton robot hip is regarded as a mass point with a distance of l meters from the hip and a mass of m , when the human body is stationary, the torque of the hip supporting the upper half of the robot is T=mlsin(θ) , the motor provides a torque K in proportion to u=KT , the value of K can be adjusted according to actual needs in different states.

[0062] 4. Software system function

[0063] The key controller design includes motor control (involving CAN address and baud rate setting, CAN communication protocol, etc.), main control board, IMU RS485 communication protocol (address and baud rate setting, inclination RS485 communication protocol, etc.), torque sensor RS485 communication protocol (write address, clear, read torque value, etc. Instructions and corresponding data format), encoder RS485 communication protocol (set baud rate, communication address, angle value clear, read angle value, etc. Instructions and corresponding data format).

[0064] The control algorithm design covers algorithm logic (state machine cases and control logic), control algorithm (squatting assist algorithm and walking assist algorithm), and algorithm programming (consisting of serial initialization, sensor transceiving, control, etc. modules, and the control part includes data processing, state recognition, control strategy and motor control module).

[0065] Advantages of the present application

[0066] ① Through the fusion of multiple sensors such as IMU, encoder and torque sensor, the wearer's body posture, joint stress and motion state information can be accurately and comprehensively obtained, so as to accurately identify the human motion state and provide appropriate assistance in time.

[0067] ② Through multi-sensor fusion and appropriate control algorithm, the exoskeleton robot can judge the real-time motion state of the human body and adjust the assist force and direction, ensure the compliance of human-machine cooperation, and reduce the discomfort in the human-machine interaction process.

[0068] ③ By running the corresponding assist algorithm for different motion states such as squatting and walking, the control accuracy of the exoskeleton is significantly improved, so that the exoskeleton robot can cope with complex and variable working environments, and effectively reduce the risk of muscle fatigue and injury of the wearer. BRIEF DESCRIPTION OF DRAWINGS

[0069] Figure 1 It is a whole mechanical structure diagram of exoskeleton

[0070] Figure 2 It is a flow chart of control method

[0071] Figure 3 It is a state identification flow chart DETAILED DESCRIPTION

[0072] The present application relates to a hip assist exoskeleton control method based on force-position information fusion, aiming to fuse multiple sensor data, accurately determine the motion state of the wearer, and then provide adaptive assistance, strengthen the wearer's motion performance and improve work efficiency. The following is a detailed explanation of the specific implementation of the present application:

[0073] 1. Hardware system composition

[0074] As Figure 1As shown, the exoskeleton robot includes a back structure 1, a waist structure 2, an upper limb wearing structure 3, a hip joint motion module 4 and a leg binding mechanism 5. The back structure 1 accommodates electrical elements such as batteries, controllers and the like, provides support and assistance; the waist structure is adjustable, suitable for different body types; the upper limb wearing structure 3 adopts a combination of inner flexible material and outer rigid load-bearing structure, improving wearing comfort; the hip joint motion module 4 is composed of a power-assisted motor, an encoder, a joint connecting disc, a thigh rod and the like, realizing motion control of the hip joint; and the leg binding mechanism 5 makes the exoskeleton leg rod closely fit the human leg.

[0075] The control system includes a motion control board (CSPACE-V4), a torque sensor, an external absolute encoder, an IMU inertial sensor 6, a lithium battery and a DC-DC power converter.

[0076] S1. Multimodal information acquisition and preprocessing

[0077] IMU data processing: the body back IMU is used to measure the body pitch angle θ body and the body pitch angular velocity v body , and the positive direction is the direction of the body bending downward. Zero processing is performed, so that the IMU angle value is 0 when standing upright, and the direction of bending is the positive direction. The left and right hip angles, angular velocity values and IMU angles in the pitch direction, angular velocity values are sent to the upper computer for monitoring. At the same time, the data error function is increased, if the current value and the value of the previous two sampling points change too much (such as in the program by judging whether the angle exceeds the reasonable range or not), the variable error is set to 1, and the variable value is sent to the upper computer to monitor the danger, and the upper computer can process to cancel the error.

[0078] Encoder data processing: the absolute encoder at the two hip motors is used to measure the angle value θ right of the right hip and the angle value θ left of the left hip, and the positive direction is the direction of the thigh kicking forward. The left hip joint encoder value is inverted, so that the directions of the left and right hips lifting forward are both positive directions. Then, the encoder values of two adjacent sampling points are differentially processed, and the right hip angular velocity v right and the left hip angular velocity v left are calculated according to the sampling time.

[0079] Torque sensor data processing: the torque sensor is installed at the output end of the hip joint exoskeleton on both sides of the hip drive motor, which can obtain the human-machine interaction force between the wearer and the exoskeleton in real time, and provide a basis for accurate control of the assistance size.

[0080] S2. Motion state recognition and judgment The specific implementation method is as follows:

[0081] Threshold setting: the body angle threshold θ stand= 15; threshold of angular velocity of moving body v move = 15; threshold of angular difference of standing hip θ hip = 15.

[0082] ① Initial state state = 0: After the system is powered on, it first enters the initial state, and the state recognition module outputs a value of 0 at this time. In this state, the system performs initialization operations, including sensor calibration, motor initialization, etc. At the same time, it waits for the sensor data to stabilize, preparing for the subsequent accurate recognition of the motion state.

[0083] ② Standing / walking state state = 1: When the following conditions are met, the system determines that it enters the standing / walking state, and the state recognition module outputs a value of 1.

[0084] The body is straight, that is, θ body < θ stand ;

[0085] The back stops moving and the hip stops moving, that is, v body < v move and θ body < θ stand .

[0086] ③ Squatting state state = 2: When θ body > 10, v body > v move , v hip_product > 50, v left > 15, v right > 15, the system determines that it enters the squatting state, and the state recognition module outputs a value of 2.

[0087] In this state, the squatting action mainly relies on gravity, but the system will continue to monitor the sensor data to prepare for the subsequent possible assistance action. At the same time, according to the mechanical characteristics in the squatting process, the timing and size of the assistance that may be needed are preliminarily estimated, for example, when reaching the squatting limit position, a certain buffer assistance is prepared.

[0088] ④ Squatting stop state state = 3: When the upper body stops bending during squatting, that is, v body < 15 and v hip_product ≤ 400, the system determines that it enters the squatting stop state, and the state recognition module outputs a value of 3.

[0089] In this state, although the body temporarily stops the squatting action, the system still monitors the sensor data, because at this time the exoskeleton may need to provide a certain support force to maintain the stability of the body in the squatting stop position, preventing instability caused by excessive squatting or sudden standing up.

[0090] State = 4: When the body is rising up, i.e. both conditions v body <-13 or v hip_product >450, v right <-5 or v left <-5, the system determines that it enters the rising state, and the state recognition module outputs a value of 4.

[0091] The rising process needs assistance, and in this state, the system starts the rising assistance algorithm and provides upward assistance to the wearer according to the preset control strategy. At the same time, according to the speed and angle change of the body rising, the assistance size is adjusted in real time to ensure that the assistance effect matches the movement needs of the wearer, for example, providing a larger assistance at the initial stage of rising to overcome gravity, and gradually reducing the assistance as the body gradually stands up.

[0092] When the upper body stops lifting, i.e. v body >v move or v hip_product <v e , the system determines that it enters the rising stop state, and the state recognition module outputs a value of 5.

[0093] At this time, although the rising action is temporarily stopped, the system still needs to maintain a certain assistance or support to prevent the body from suddenly sinking or swaying due to the loss of assistance. At the same time, the system determines whether further action adjustment is needed according to the current posture and sensor data, such as entering the rising state again or switching to other states.

[0094] Error state state = 0: When the left and right hip angles exceed the range (less than 40 or greater than 150), or the body angle exceeds the range (less than 40 or greater than 150), or error is 1, switch to the initial state.

[0095] S3, assistance strategy implementation:

[0096] ① Standing / walking state assistance: In the standing / walking state (move = 1), the assistance strategy mainly includes reducing friction and the impact of leg bar gravity on wearer movement, and compensating for hip torque according to the current angle and speed.

[0097] The zero-force dragging algorithm uses a zero-force control strategy based on torque compensation, which enables the robot driver to operate in current / torque mode. This strategy precisely controls the servo motor to provide the gravity and friction compensation required for robot movement. The control torque , where is the gravity term in the robot dynamics equation, which can be simplified in the exoskeleton robot to a function of the trigonometric function θ of the joint rotation angle sin(θ) , i.e. τG =Ksin(θ) ; friction torque at the joint (composed of coulomb friction and viscous friction). Thus the control torque of the knee and hip joint in swing phase can be obtained τ knee and τ hip , by adjusting the controller parameters K i (i=1,2…7) (can be estimated by mass, center of mass, and tested by adjusting parameters in real machine), to achieve more flexible human-computer interaction and offset the gravity.

[0098] Based on the torque compensation algorithm of the hip joint speed, two sampling interval intervals are set. The large sampling interval (usually more than 10 sampling points) is used to judge the acceleration or deceleration process of the hip joint, and the direction of the control torque is determined; the small sampling interval (usually less than 5 sampling points) is used to calculate the value of torque compensation, to enhance the following of the robot. The specific implementation can determine the calculation method of the compensation torque according to the monitoring and calculation method of the speed and angle change in the program, for example, according to the speed data v right , v left , etc. to judge acceleration or deceleration, and according to the angle data θ right , θ left , etc. to calculate the compensation value.

[0099] ②Squatting state assistance (preparation stage): in the squatting state (move=2), although mainly relying on gravity squatting, the system will estimate the force condition in the squatting process according to the current squatting speed (v body , v right , v left , etc.), angle (θ body , θ right , θ left , etc.) and the preset mechanical model.

[0100] When approaching the squatting limit position (which can be judged by angle data such as θ body approaches a certain large value), the system is ready to start the buffer assistance to prevent discomfort or injury risk caused by excessive gravity. At this time, the system starts to adjust the output torque of the motor, so that it can provide a certain reverse resistance at the right time, slow down the squatting speed, and realize smooth squatting.

[0101] ③Squatting stop state assistance (support stage): in the squatting stop state (move=3), the system provides support force to maintain the stability of the body in the squatting stop position.

[0102] The part above the hip of the exoskeleton robot is regarded as a mass of l meters from the hip mThe torque of the hip support robot upper body when the human body is static is T=mlsin(θ) (wherein θ is the back pitch angle), the motor should provide a corresponding proportional support force. If a torque with a proportional coefficient of K is provided, then u=KT , in this way, the wearer is provided with stable support, and the burden on the leg muscles is reduced.

[0103] (4) Standing up state assistance (main assistance stage): the standing up assistance adopts PD control based on position error. The control law is:

[0104]

[0105] wherein e(t)=θ d - θ is the error θ d is the desired angle, θ is the real-time measured motor angle value, K P is the proportional coefficient, K D is the differential coefficient. When the standing up assistance is performed, the desired angle can be set to 0.

[0106] The system calculates the current position error according to the real-time hip angle (θ right , θ left ) and angular velocity (v right , v left ) data, and provides a suitable assistance torque by adjusting the proportional coefficient K P and the differential coefficient K D . In the initial stage of standing up, the position error is large, and the system provides a large assistance torque to overcome gravity and start the standing up action; as the body gradually approaches the upright state, the position error decreases, and the assistance torque also decreases accordingly, until the upright position is reached, the assistance torque gradually decreases to zero, and a smooth transition is achieved.

[0107] (5) Standing up stop state assistance (stable holding stage): in the standing up stop state (move=5), the system continues to provide a certain support force to ensure the stability of the body in the upright position.

[0108] Similarly, the support force calculation method similar to the squatting stop state is adopted, and the current body posture (θ body , θ right , θ leftThe required support torque is calculated based on the relationship between the sensor data (such as θ

[0109] S3. Motor Control:

[0110] ① Motor Initialization: The motor control module completes the motor initialization operation within 3 seconds after the system is powered on. The left and right hip motors are turned on, and the corresponding instructions are sent to set the speed limit of the motors. These initialization settings ensure that the motors can work stably and safely in the subsequent running process, and can accurately output the required torque and speed according to the control strategy.

[0111] ② Send Current Value: After 3 seconds of power-on, the motor control module mainly performs the operation of sending the current value. After converting the control torque value calculated by the control strategy module into current, it is sent to the driver. The driver controls the left and right hip assist motors to run in the current loop according to the received current value, so that the torque output by the motor matches the control torque value calculated, thereby realizing accurate control of the exoskeleton and providing appropriate assistance for the wearer. In the whole process, the motor control module constantly updates the sent current value according to the sensor data (such as θ right , θ left , v right , v left , θ body , v body , etc.) and the adjustment of the control strategy, to ensure that the assistance effect of the exoskeleton is always consistent with the movement needs of the wearer.

Claims

1. A hip-assisted exoskeleton control method based on force-position information fusion, protecting the following steps: S1. Multimodal Information Acquisition and Preprocessing The system utilizes inertial sensors mounted on the back to collect body pitch angle and pitch velocity information, an absolute encoder to collect hip joint angle information, and a torque sensor to collect hip joint torque information. The collected inertial sensor data is filtered to remove noise interference, the absolute encoder data is calibrated to ensure the accuracy of angle measurements, and the torque sensor data undergoes zero-drift correction. S2. Motion State Recognition and Judgment The body pitch angle and angular velocity are collected by inertial sensors and combined with the angle value of the absolute encoder. Based on the preset threshold conditions, the state machine switches between states, including standing / walking, squatting, squatting and stopping, standing up and standing and stopping. Set the body pitch angle threshold θ stand Body pitch angular velocity threshold v move Hip position difference threshold θ e Hip angular velocity product threshold v e Movement states include standing / walking, squatting, squatting still, standing up, and standing still. The specific determination methods are as follows: When the upper body bends greater than θ stand And the bending angular velocity reaches v move And the encoder position difference is less than θ e And the product of the left and right angular velocities is greater than v. e When switching from a standing / walking position to a squatting position, When the upper body stops bending during the squat, that is, when the upper body angular velocity is less than v. move And the product of the left and right angular velocities is less than v. e At that time, enter the squatting stop state. When the body stops squatting or standing up, if the body rises upwards, that is, the angular velocity of the upper body is less than v. move or the product of angular velocities is greater than v e Then you will enter the standing position. When the upper body stops lifting during the rising process, that is, when the angular velocity of the upper body is greater than v. move Or the product of angular velocities is less than v e At that time, enter the standing still state. When the body continues to rise from the stopped position, and the above conditions for rising are met, the rising position is maintained; when the body angle is less than θ stand And the upper body angular velocity is less than v move When switching from a standing position to a standing / walking position, the system collects real-time data on pitch angle and angular velocity from the user's back inertial sensor, as well as angle data from the absolute encoder, to identify and determine the motion state and employ different assist strategies in different motion states. The control algorithms and execution details for different states are as follows: ① If the user is standing / walking, the walking assistance algorithm is executed, including a zero-force drag algorithm to calculate the torque against gravity and friction, and a torque compensation algorithm based on hip joint velocity to calculate the compensation torque. The two are added together to obtain the final control torque. ② If the person is standing up, calculate the assist torque according to the PD control law for standing assistance, and calculate the final control torque in combination with the support force compensation. ③ If the user is in a squatting or standing position, only the control torque of the support force compensation portion is calculated, including two control modes: Follow-up control mode: Based on the torque changes sensed by the left torque sensor and / or right torque sensor caused by the user's instantaneous movements, the movement trend of the user's left and / or right hips is determined, and the left and / or right hip joint assist mechanism is controlled to move accordingly. II. Left and right synchronous control mode: Based on real-time measurements of angle, angular velocity, angular acceleration, and assist torque detected by the backpack inertial sensor, the user's torso posture is determined. Then, combined with measurements of the angles and changes in the left and right hip joint assist mechanisms from the absolute encoders, the user's current movement type is inferred. Correspondingly, the left and right hip joint assist mechanisms are controlled to perform the appropriate actions to provide assistance. Detailed judgment is as follows: If the inertial sensor on the back detects that the user's upper body is basically in an upright position, the left and right hip joint assist mechanisms alternately bend / extend to determine that the user is in a walking state. The angle, amplitude, and speed of the hip joint assist mechanisms on the left and right sides of the user are the same, but the directions are opposite. If the inertial sensor on the back detects that the user's upper body is leaning forward, the left and right hip joint assist mechanisms extend simultaneously, indicating that the user is in a standing position. The angle, amplitude, speed, and direction of the movement of the left and right hip joint assist mechanisms are the same. If the inertial sensor detects that the user's upper body is leaning to the left, the left hip joint assist mechanism bends and the right hip joint assist mechanism extends, indicating that the user is turning to the left; the left and right hip joint assist mechanisms move in opposite directions, and the angle, amplitude, and speed of the user's left hip joint assist mechanism are less than those of the user's right hip joint assist mechanism. If the inertial sensor detects that the user's upper body is leaning to the right, the right hip joint assist mechanism bends and the left hip joint assist mechanism extends, indicating that the user is turning to the right; the user's left and right hip joint assist mechanisms move in opposite directions, and the angle, amplitude, and speed of the user's right hip joint assist mechanism are less than those of the user's left hip joint assist mechanism. If the back IMU sensor detects that the user is leaning forward, the hip angle increases and the upper body flexion angular velocity reaches a set threshold v. move Meanwhile, the angle difference measured by the absolute encoders at the two hip motors is less than the threshold θ. e And the product of the left and right angular velocities is greater than the threshold v. e It determines when the user starts to squat; When in a squatting or standing position, if the back IMU sensor detects that the user's upper body begins to lean back, it determines that the user has entered the standing position. During the standing up process, the upper body leans backward continuously, and the hip angle gradually decreases. The system uses PD control based on position error to calculate the assist torque and combines it with support force compensation to provide upward assistance to the user. The assist torque is calculated based on real-time hip angle and angular velocity data. It provides greater assistance to overcome gravity in the early stage of standing up, and the assist torque gradually decreases as the body gradually straightens. When the upper body stops lifting and enters the standing up stop state, the system continues to provide support force to ensure that the body is stable in the upright position, while continuously monitoring sensor data to determine whether further movement adjustments are needed.

2. The hip-assisted exoskeleton control method based on force-position information fusion according to claim 1, characterized in that, The entire control process also includes an error handling mechanism for sensor data jumps, specifically: The data processing module processes the sensor values, including inverting the left hip joint encoder values ​​to unify the left and right directions, performing differential processing on the encoder values ​​of adjacent sampling points and calculating the hip angular velocity, and zeroing the IMU. Additionally, if the current value differs significantly from the values ​​of the previous two sampling points, the variable `error` is set to 1 to indicate a data jump. The status recognition module determines the motion status based on sensor data. When the error is 1, the status is switched to the initial 0, the assist motor stops moving, and the wearer's safety is ensured.

3. The hip-assisted exoskeleton control method based on force-position information fusion according to claim 1, characterized in that, The motor control module operation includes motor initialization and sending current values, specifically: The initialization of the assist motors is completed within 3 seconds of the exoskeleton robot being powered on. This includes enabling the left and right hip assist motors and sending commands to set the motor speed limits. Three seconds after power-on, the motor control module only performs the operation of sending current value. It converts the control torque value calculated by the control strategy module into current and sends it to the driver, so that the assist motor runs according to the control torque under the current loop. This enables the exoskeleton to assist the human hip, helping the human to complete bending, standing, and walking movements, reducing the burden on the human body, and reducing the risk of muscle fatigue or injury. At the same time, it continuously monitors sensor data and adjusts the control strategy and torque output in real time according to the status changes.

Images