Reconfigurable exoskeleton assisting robot and control method, device and storage medium

Abstract

The application discloses a reconfigurable exoskeleton assisting robot and a control method, equipment and storage medium thereof, and the control method comprises the following steps: collecting human motion data in real time through a sensor module; calculating Euler angles by using a quaternion fusion algorithm; predicting a target position based on Kalman filtering; compensating tracking errors through a PID controller; and switching working modes according to motion states. The application breaks through the single interaction limitation, takes into account the operation convenience and function expandability, and realizes efficient human-machine cooperation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of manufacturing, in particular to a reconfigurable exoskeleton assisting robot and a control method, device and storage medium. BACKGROUND

[0002] With the continuous increase of labor intensity in the fields of industrial manufacturing, logistics transportation and construction, industrial workers are long-term exposed to the risk of musculoskeletal diseases caused by high-intensity repetitive work. As human function auxiliary equipment, the reconfigurable assisting system can effectively alleviate labor injury, but the existing technology still has significant defects: (1) functional limitation: most of them focus on single assistance (such as upper limb lifting or lower limb walking), lack of whole body collaborative assistance ability, and cannot adapt to complex scenes (such as lifting and walking at the same time). (2) Rigid structure and low adaptability: the traditional reconfigurable assisting system adopts fixed mechanical structure, which is difficult to adapt to different body types of users or dynamic task requirements, and the human-machine motion coupling is poor, which is easy to cause secondary injury. (3) Weight and performance contradiction: in order to ensure the structural strength, the mainstream products adopt metal materials (such as aluminum alloy), which leads to the weight of the whole machine exceeding 15 kg, increasing the burden of users; lightweight design often sacrifices the power output. (4) Intention recognition lag: the motion perception system relying on a single sensor has low prediction accuracy and high delay in complex postures, causing lag in assisting response. (5) Insufficient interactive experience: the control relies on physical keys or handheld devices, which cannot free the user's hands, and lacks multi-modal interactive channels. SUMMARY

[0003] The purpose of the present application is to provide a reconfigurable exoskeleton assisting robot and a control method, device and storage medium, to design a double-channel efficient human-computer interaction system, to propose a motion intention prediction algorithm based on distributed multi-IMU fusion, to effectively improve the real-time performance and accuracy of human motion state perception in complex postures, to improve the traditional intention recognition lag, to innovate the voice and APP double-channel control scheme, to break through the single interaction limitation, to consider the operation convenience and functional expandability, to realize efficient human-computer collaboration, to enhance the interactive experience, and to free the user's hands, so as to solve the functional limitation of the existing technology, to realize the flexible reconfiguration of exoskeleton function and load, and to flexibly switch to single-arm, double-arm, lower limb and whole body four working modes according to the use scene.

[0004] Technical scheme: the control method of the reconfigurable exoskeleton assisting robot comprises the following steps:

[0005] S1: collecting human motion data in real time through a sensor module;

[0006] S2: calculating Euler angles by using a quaternion fusion algorithm; including calculating an initial quaternion; updating an iterative quaternion; correcting the magnetic fusion and yaw drift; converting the updated quaternion into roll angle, pitch angle and yaw angle output;

[0007] S3 predicting target position based on Kalman filter; including constructing state equation and observation equation; calculating predicted state covariance matrix; updating state estimation and covariance using Kalman gain;

[0008] S4 compensating tracking error through PID controller;

[0009] S5 switching working mode according to motion state.

[0010] Further, step S1 is specifically as follows: using distributed nine-axis IMU network to collect roll angle, pitch angle, heading angle data and three-axis acceleration data of multiple parts of human body in real time; and performing sensor initialization, data reading and preprocessing according to the collected data.

[0011] Further, in step S2, the quaternion fusion algorithm includes the following steps:

[0012] S21 calculating initial quaternion: calculating initial attitude angle based on accelerometer and magnetometer data and converting the initial attitude angle into initial quaternion; wherein the attitude angle includes roll angle, pitch angle and yaw angle;

[0013] S22 updating quaternion iteration: using Mahony complementary filtering algorithm to fuse accelerometer and gyroscope data to suppress drift; including acceleration normalization, attitude error calculation and PI controller correction;

[0014] S23 correcting magnetometer fusion and yaw drift: using magnetometer data to provide absolute heading reference, converting normalized magnetic field vector to navigation coordinate system, calculating reference heading angle to correct gyroscope integral drift;

[0015] S24 outputting Euler angle: converting the updated quaternion into roll angle, pitch angle and yaw angle for output, the formula is as follows:

[0016] ;

[0017] ;

[0018] ;

[0019] wherein, is the roll angle, is the pitch angle, is the yaw angle; is the quaternion.

[0020] Further, in step S3, Kalman filter prediction includes the following steps:

[0021] S31 constructing state equation and observation equation:

[0022] ;

[0023] ;

[0024] wherein, is the input signal, is the state variable, is the process noise, is a state transition model acting on , is an input control model acting on the controller vector , is the observation variable, is the observation noise, is an observation model;

[0025] S32 prediction step: compute the predicted state covariance matrix, formula is:

[0026] ;

[0027] wherein, denotes the covariance of the predicted state estimate (measures the prediction uncertainty), is the process noise covariance matrix,

[0028] S33 update: update the state estimate and covariance using the Kalman gain, formula is:

[0029] ;

[0030] ;

[0031] ;

[0032] wherein, denotes the Kalman gain, is the covariance matrix of the state estimate at time step k based on the estimate at the previous time step k-1. It represents the uncertainty of the state estimate at time step k-1 on the state estimate at time step , is the observation matrix, is the covariance matrix of the observation noise, is the state estimate update, is the observation value at time step , is the covariance update, is the identity matrix.

[0033] Further, in step S4, the PID controller compensation comprises the following steps:

[0034] S41 defines error variable :

[0035]

[0036] wherein is the motor predicted position, is the motor actual position;

[0037] S42 calculates compensation amount Δq(t):

[0038] ;

[0039] wherein, the proportional term P is , used to quickly respond to the current error rate of change, and the Kalman gain forms a "prediction-proportional" double-loop regulation; the integral term I is , used to accumulate historical errors to eliminate the steady-state deviation caused by mechanical static friction; the differential term D is , used to predict the error change trend to suppress the overshoot when the arm swings quickly;

[0040] S43 parameter setting: Ziegler-Nichols critical proportional degree method is used to set the PID parameters.

[0041] Further, step S43 is specifically as follows: gradually increase the proportional gain until the system is dithered, and then calculate the final parameters based on the critical gain , / , .

[0042] Further, step S5 is specifically as follows: based on the fused Euler angle data and three-axis acceleration values, and combined with the preset motor rotation range threshold, the weights of the three angles are adaptively distributed to identify the current upper limb and lower limb state of the human body, so as to switch the upper limb following mode, the upper limb lifting mode, the lower limb walking mode and the human body carrying mode.

[0043] The reconfigurable exoskeleton assisting robot described in the application is realized by using any one of the control methods, comprising: a backpack back plate provided with a quick release locking seat; a battery module and a main control module installed on the front face of the backpack back plate; a detachable joint module connected with the backpack back plate through the quick release locking seat; a sensor module distributed on the detachable joint module and the backpack back plate; a voice module installed on the back face of the backpack back plate for receiving voice instructions; wherein the main control module controls the joint module movement in real time through the sensor module.

[0044] An electronic device according to the application comprises a memory and a processor, the memory storing a computer program, the processor implementing the steps of any of the methods when executing the program.

[0045] A computer-readable storage medium according to the application stores a computer program, the program implementing the steps of any of the methods when executed by a processor.

[0046] Beneficial effects: Compared with the prior art, the application has the following significant advantages: the fast disassembly type reconfigurable joint and the load distribution structure design idea are adopted, the flexible reconfiguration of the load of the power assisting system is realized, four working modes can be flexibly switched according to the use scene, the motion intention prediction algorithm based on distributed multi-IMU fusion is proposed, the real-time performance and the accuracy of the human motion state perception under a complex posture are effectively improved, the double-channel efficient human-computer interaction system is designed, the voice and APP double-channel control scheme is innovated, the single interaction limitation is broken through, the operation convenience and the function expansibility are taken into account, and efficient human-computer collaboration is realized. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 is a structural schematic diagram of a reconfigurable exoskeleton assisting robot according to the application (reverse side);

[0048] Figure 2 is a structural schematic diagram of a reconfigurable exoskeleton assisting robot according to the application (front side);

[0049] Figure 3 is a structural exploded view of a master control module according to the application;

[0050] Figure 4 is a structural exploded view of a battery module according to the application;

[0051] Figure 5 is a nine-axis IMU data acquisition flowchart according to the application;

[0052] Figure 6 is a Kalman filtering and PID optimization collaborative control algorithm flowchart;

[0053] Figure 7 is a motor control flowchart. DETAILED DESCRIPTION

[0054] The technical solutions of the application will be further described below with reference to the drawings.

[0055] As shown in Figure 5 , the embodiment of the application further provides a control method of a reconfigurable exoskeleton assisting robot, a Kalman filtering and PID optimization collaborative control algorithm is combined to drive the multi-joint motors to work cooperatively, and convenient human-computer collaborative assistance is realized through voice interaction. The method comprises the following steps:

[0056] Step 1: Data collection using a nine-axis IMU;

[0057] Step 2: Data fusion: to generate stable and reliable Euler angle data. Including the following steps:

[0058] (21) Initial quaternion calculation

[0059] The present application first converts the collected sensor data (accelerometer , magnetometer ) into Euler angles, and calculates the initial quaternion based on this. Specifically, the initial attitude angle (roll angle , pitch angle , yaw angle ) is calculated by the present application using the following formula:

[0060] (1);

[0061] (2);

[0062] (3);

[0063] Where ; the present application calculates the initial quaternion [] from the above Euler angles as follows:

[0064] (4);

[0065] (5);

[0066] (6);

[0067] (7);

[0068] In formulas (4)-(7) , , , , and are as follows:

[0069] (8);

[0070] (9);

[0071] (10);

[0072] (11);

[0073] (12);

[0074] (13);

[0075] (22) quaternion iterative update: the Mahony complementary filter algorithm is used to iteratively update the quaternion, and the accelerometer and gyroscope data are fused to suppress drift, specifically including:

[0076] Acceleration correction: first, the accelerometer and gyroscope are complementarily corrected, and the acceleration is normalized, as shown in the following formula:

[0077] (14);

[0078] Next, the present application calculates the cross product of the accelerometer gravity vector and the gyroscope predicted gravity vector to generate an attitude error signal. To effectively fuse and suppress drift, the present application is provided with a PI controller (proportional gain KP=1.5, integral gain KI=0.005). The controller uses the attitude error signal to quickly respond to the current error, real-time corrects the instantaneous deviation of the gyroscope, and effectively suppresses the cumulative historical error through the integral term, and finally outputs the corrected angular velocity gyro_correct.

[0079] Quaternion update: the present application uses the first-order Runge-Kutta method with less calculation to update the quaternion, which only needs to use the corrected gyroscope angular velocity , and the formula is as follows:

[0080] (15);

[0081] where is the quaternion, is the gyroscope angular velocity.

[0082] Step 3: magnetometer fusion and yaw drift correction: to overcome the problem of significant drift of the gyroscope in the yaw angle, the present application introduces the magnetometer data to provide an absolute heading reference. The magnetometer measures the original magnetic field vector in the body coordinate system The data processing flow of the magnetometer of the present application is as follows:

[0083] (16);

[0084] Since the carrier attitude will affect the magnetometer measurement value, the present application uses the current attitude quaternion to calculate the rotation matrix , which converts the normalized magnetic field vector to the navigation coordinate system, as shown in the following formula:

[0085] (17);

[0086] where the rotation matrix is the rotation matrix from the body frame to the navigation frame, which can be calculated from the quaternion, as shown in the following formula:

[0087] (18);

[0088] Reference heading calculation and drift correction: the present application calculates the reference heading angle based on the converted horizontal components and where and are the horizontal components of the magnetic field in the navigation frame. The attitude is then calculated by integrating the angular velocity from the gyroscope, but there is a drift problem, especially the yaw angle. The absolute heading information provided by the magnetometer can be used to correct this drift. Finally, the updated quaternion is converted to Euler angles.

[0089] The absolute heading information is used by the present application to correct the yaw angle drift generated by the gyroscope integration, significantly improving the long-term stability of the yaw angle.

[0090] Final attitude output: after completing the fusion update, the present application converts the quaternion to Euler angles for output:

[0091] (19);

[0092] (20);

[0093] (21);

[0094] where is the roll angle, is the pitch angle, is the yaw angle.

[0095] Through the data fusion algorithm of the present application, high stability and reliability of the Euler angle data are achieved, providing an accurate attitude reference for high-precision control of the motor.

[0096] Step 4: cooperative control algorithm based on Kalman filtering and PID optimization: the present application uses the above-mentioned nine-axis gyroscope to collect the Euler angle data of the current body five position sensors, obtaining the roll angle, pitch angle, heading angle data of the five parts of the human body, as well as the acceleration data in the x, y, z directions. The two nine-axis gyroscopes of the upper limbs control the brushless motor of the upper limbs, and the three nine-axis gyroscopes of the back and lower limbs control the brushless motor of the lower limbs.

[0097] Kalman prediction (state estimation):

[0098] ​The Kalman filtering algorithm is used to predict the target rotating position of the motor.

[0099] Algorithm construction: the state equation and the observation equation of the system are constructed, and the initial state is set, and the formula is as follows:

[0100] (22);

[0101] (23);

[0102] Wherein is an input signal, is a state variable, is process noise, is a state transformation model acting on is an input control model acting on the controller vector is an observation variable, is observation noise, is an observation model. Prediction step:

[0103] In order to estimate the error, the covariance error matrix needs to be calculated to measure the accuracy of the estimated value. The formula is as follows:

[0104] (24);

[0105] Wherein indicates the covariance of the predicted state estimation (measuring the prediction uncertainty),

[0106] is the process noise covariance matrix. Update step:

[0107] According to the observation value and the predicted value at the current time, the state estimation value at the current time is calculated. This estimation value is a more accurate estimation value, because it has considered the observation value at the current time. The error covariance matrix of the state estimation value is calculated by the error covariance matrix calculated in the prediction step, the observation noise covariance matrix and the Kalman gain. The update formula is as follows:

[0108] (25);

[0109] (26);

[0110]

[0111] ​​ (27);

[0112] From the above formula, the Kalman gain Actually represents the proportion of the model prediction error (Predicted error) and the measurement error (Measurement error) in the process of optimal state estimation (as follows), that is When , that is, the prediction error is 0, and the state value of the system depends entirely on the predicted value; and when , that is, the measurement error is 0, and the state value of the system depends entirely on the observation value.

[0113] PID real-time feedback compensation:

[0114] On the basis of Kalman prediction, the application introduces a PID controller to form a closed-loop feedback, and realizes accurate tracking of the motor position through real-time error correction. Its control logic and the "prediction-correction" mechanism of Kalman filtering form a complement: Kalman filtering mainly deals with system model uncertainty (such as sensor noise, modeling error), and PID focuses on suppressing real-time tracking errors caused by external disturbances, load changes, mechanical static friction, etc.

[0115] The application defines an error variable for PID calculation, and the formula is as follows:

[0116] (28);

[0117] Wherein is the predicted position of the motor, is the actual position of the motor. The purpose is to calculate the compensation amount by the PID algorithm described below, which is used to suppress Kalman prediction, and the formula is as follows:

[0118] (29);

[0119] Wherein the proportional term P is , which is used to quickly respond to the current error change rate and form a "prediction-proportional" double-loop regulation with the Kalman gain; the integral term I is , which is used to accumulate historical errors and eliminate steady-state deviations caused by mechanical static friction; and the differential term D is , which is used to predict the error change trend and suppress overshoot when the arm swings quickly.

[0120] Parameter setting: the application adopts Ziegler-Nichols critical proportional degree method to preliminarily set PID parameters, and the action of and is closed, only the action of proportional term is reserved, starting from a small 1.0, gradually increasing, observing the response of the system. Gradually increase until the motor output starts to produce obvious jitter. Then according to the empirical formula of Ziegler-Nichols , and Calculate the PID parameters suitable for the system, , 10ms, and finally adjust according to the actual effect of the motor, then calculate the final control signal by weighting the predicted value and the compensation value of the motor, and input the motor driver to drive the brushless motor to rotate to the target position accurately. The flow chart of the Kalman filter and PID optimization cooperative control algorithm is shown in Figure 6 .

[0121] The Kalman filter and PID joint control algorithm of the application overcomes the limitations of single control method through clear division of responsibilities (KF predicts global optimum, and PID compensates local error) and efficient prediction-compensation closed-loop cooperative mechanism. It not only significantly improves the accuracy and speed of motor position tracking, but more importantly, in the face of sensor noise, model uncertainty, load changes, mechanical friction and other disturbance sources, it shows excellent stability and robustness, providing a reliable, smooth and rapid response execution control core for high dynamic applications such as exoskeletons.

[0122] Step 5: Based on the fused Euler angle data and three-axis acceleration value, the application combines the preset motor rotation range threshold to judge the adaptive distribution of the weight of the three angles to identify the current state of the upper limbs and lower limbs of the human body, so as to switch the upper limb following mode, upper limb lifting mode, lower limb walking mode and human body carrying mode. The specific implementation process is as follows:

[0123] The zero value of the motor is set according to the initial standing arm vertical state of the human body after the main control is turned on, the weighted value of the upper limb gyroscope is the base point of the motor rotation, the weighted value of the gyroscope is matched with the rotation value of the motor according to the maximum amplitude of the actual human body limb rotation, and finally the motion state is judged according to the value of three-axis acceleration, wherein the motion mode includes upper limb following mode, upper limb lifting mode, lower limb walking mode, lower limb standing mode and human body carrying mode, and the details are as follows:

[0124] First, the host is powered on, and the motor is in the upper limb following mode and the lower limb walking mode, and the three-axis acceleration values of the gyroscope are recorded every 10 ms. When the latest acceleration value is greater than the last recorded value, the upper limb following mode and the lower limb walking mode of the motor are switched to the limb prediction mode, so that the motor drives the human body to move in the predicted direction. The application processes the data of the five gyroscopes processed by the host, judges the current state of the human body in turn to change the control mode of the motor, and also has a force unloading function. When the current of the motor exceeds the threshold (20A), the human body is protected to enter the force unloading state. The upper limb following mode and the lower limb walking mode:

[0125] When the exoskeleton assisting robot is powered on, the upper limbs are in the upper limb following mode by default, and the lower limbs are in the lower limb walking mode by default. In this mode, the distributed nine-axis IMU network Real-time capture of human motion intention, Kalman filter and PID optimization algorithm driven by ESP32 host + STM32 coprocessor running 4 brush joint motors (peak torque 12N·m), accurate prediction of user intention, 10ms level low delay cooperative assistance, ensure that the exoskeleton assisting effect is consistent with the motion demand of the pedestrian.

[0126] Upper limb lifting mode:

[0127] When the user lifts the load to a certain position in the upper limb following mode, the mode can be modified to the upper limb lifting mode through the voice control module or the mobile phone APP. In this mode, the position proportional gain KP and the speed differential gain KD are simultaneously increased, and the single arm can stably provide a peak assistance effect of 60N, and the double arms can provide a peak assistance effect of 120N, to help maintain the current position of the upper limbs and reduce the burden of the upper limbs. It is suitable for use in scenarios such as photographers or workers who engage in lifting work for a long time

[0128] Human carrying mode:

[0129] When the gyroscope on the back detects that the body bending angle reaches the set bending angle threshold, the gyroscope values of the upper limbs and the lower limbs are detected Whether to reach the bending state At this time, the three-axis acceleration of the upper and lower limbs is judged And Greater than the acceleration value of the last time , When the lower limbs motor enters the human body carrying mode, the synchronous lifting position proportional gain KP and the speed differential gain KD are increased, the single leg can stably provide a peak assistance effect of 85N, the single arm can stably provide a peak assistance effect of 60N, to help carry heavy objects and reduce the burden on the whole body, until the human body returns to the normal standing state, and then return to the following mode. It is suitable for use in scenarios such as carrying heavy objects for a long time, such as carrying workers. The overall motor control flow chart is shown in Figure 7 .

[0130] As shown in Figures 1-2 , the embodiment of the application provides a reconfigurable exoskeleton assistance robot, which can be flexibly switched according to the use scene and realized reconfiguration; the robot comprises a backpack 108, a battery module 104, a main control module 105, a voice module 202, a sensor module (including a first sensor 102, a second sensor 107, a third sensor 203, and a fourth sensor 204), and a detachable joint module (including a left upper joint 101, a right upper joint 106, a left lower joint 103, and a right lower joint 109); the first sensor 102 and the second sensor 107 are connected with the left upper joint 101 and the right upper joint 106 respectively; the third sensor 203 and the fourth sensor 204 are connected with the left lower joint 103 and the right lower joint 109 respectively. The first sensor 102, the third sensor 203, the third sensor 203, and the fourth sensor 204 are internally integrated with a gyroscope and a geomagnetic sensor.

[0131] The backpack 108 is responsible for installing the battery module 104, the main control module 105, the left upper joint 101, the right upper joint 106, the left lower joint 103, and the right lower joint 109. The backpack 108 comprises three base plates, two shoulder joint module locking seats, and two leg joint module locking seats, which are responsible for the installation and adjustment of the corresponding joint modules.

[0132] The connecting device of the left upper joint 101, the right upper joint 106, the left lower joint 103, and the right lower joint 109 and the backpack 108 is a quick-release locking seat. When installing the corresponding joint, the joint is inserted into the quick-release locking seat. The upper end of the quick-release locking seat adopts a slotted locking pipe matched with a screw type locking terminal. After the wearer inserts the joint module into the locking seat and tightens the screw type locking terminal, the installation of the corresponding joint module can be quickly completed.

[0133] As shown in Figure 4 , the battery module 104 comprises, in sequence, a power module fixing module 401, a power module bottom plate 402, a lithium battery 403, and a power module upper cover 404, and is responsible for providing power supply for the whole exoskeleton robot system.

[0134] As shown in Figure 3As shown, the main control module 105 is responsible for controlling the movement of the system, in turn including: main control bottom plate 301, main control circuit board 302, main control support plate 303, screen bottom plate 304, main control upper cover plate 305, power button 306.

[0135] The voice module 202 includes a voice board, an external microphone, a voice module upper cover, and a voice module lower cover. Through the voice module, the user's voice instructions can be recognized and analyzed, and the exoskeleton can be controlled without using both hands, making it more convenient to use.

[0136] The left upper joint 101, the right upper joint 106, the left lower joint 103, and the right lower joint 109 include a joint module upper cover, a joint module base, a needle bearing, a carbon fiber tube, a carbon fiber tube locking seat, a servo motor, a servo motor output connector, an output connection carbon fiber tube, a carbon fiber tube sleeve, a binding member, and a binding member connector. The carbon fiber tube in the left upper joint 101, the right upper joint 106, the left lower joint 103, and the right lower joint 109 is installed in the joint module base through the needle bearing and is fixed through the carbon fiber tube locking seat, providing the shoulder with the freedom of abduction and adduction. The servo motor is fixed to the joint module base through screws, the joint module upper cover is fixed to the joint module base cover through screws, the servo motor output connector is connected to the servo motor through six screws, the output connection carbon fiber tube is connected to the servo motor output connector through screws, the binding member connector is connected to the output connection carbon fiber tube through screws, and the binding member is connected to the binding member connector through a suture. The first sensor 102, the second sensor 107, the third sensor 203, and the fourth sensor 204 are connected to the binding member through a suture, and the length of the binding member can be freely adjusted according to the user's body parameters.

Claims

1. A control method for a reconfigurable exoskeleton-assisted robot, characterized in that: The following steps are involved: S1 collects human motion data in real time through sensor modules; S2 uses a quaternion fusion algorithm to calculate Euler angles. This includes calculating the initial quaternion; updating the iterative quaternion; correcting the magnetometer fusion and yaw drift; and converting the updated quaternion into roll, pitch, and yaw angle outputs. The quaternion fusion algorithm includes the following steps: S21 calculates initial quaternion: calculates initial attitude angle based on accelerometer and magnetometer data and converts it into initial quaternion; the attitude angle includes roll angle, pitch angle, and yaw angle; S22 quaternion update: uses the Mahony complementary filtering algorithm to fuse accelerometer and gyroscope data to suppress drift; includes acceleration normalization, attitude error calculation, and PI controller correction; S23 Correction of Magnetometer Fusion and Yaw Drift: Using magnetometer data to provide an absolute heading reference, the normalized magnetic field vector is converted to the navigation coordinate system, and the reference heading angle is calculated to correct for gyroscope integration drift; S24 converts the updated quaternion into roll, pitch, and yaw angles, which are output as Euler angles. The formula is as follows: ； ； ； in, is the roll angle, is the pitch angle, is the yaw angle; is a quaternion; S3 predicts target position based on Kalman filtering; includes constructing state equations and observation equations; calculating the predicted state covariance matrix; and updating state estimates and covariance using Kalman gain. S4 compensates for tracking error through PID controller; S5 switches working modes according to the motion status.

2. The control method of a reconfigurable exoskeleton-assisted robot according to claim 1, characterized in that: Step S1 is as follows: using a distributed nine-axis IMU network to collect roll angle, pitch angle, heading angle data and three-axis acceleration data of multiple parts of the human body in real time; and performing sensor initialization, data reading and preprocessing based on the collected data.

3. The control method of a reconfigurable exoskeleton-assisted robot according to claim 1, characterized in that: In step S3, the Kalman filter prediction includes the following steps: S31 constructs the state equation and observation equation: ； ； in, is the input signal, is the state variable, is the process noise, For the effect The state transition model on is acting on the controller vector The input control model on is the observed variable, is the observation noise, is the observation model; S32 prediction step: Calculate the predicted state covariance matrix, the formula is: ； in, represents the covariance of the predicted state estimate, is the process noise covariance matrix, S33 update: Use Kalman gain to update state estimation and covariance, the formula is: ； ； ； in, represents the Kalman gain, This is the estimated value of the covariance matrix of the state estimate at time step k based on the previous time step k-1, which means the estimated value of the covariance matrix of the state estimate at time step k-1 for time step The uncertainty of the state estimate, is the observation matrix, is the covariance matrix of the observation noise, is the state estimate update, is the time step The observed value of is the covariance update, is the identity matrix.

4. The control method of a reconfigurable exoskeleton-assisted robot according to claim 1, characterized in that: In step S4, PID controller compensation includes the following steps: S41 defines the error variable : in Predict the position for the motor, is the actual position of the motor; S42 calculates the compensation amount Δq(t): ； Among them, the proportional term P is , which is used to quickly respond to the current error change rate and form a "prediction-proportional" dual-loop regulation with the Kalman gain; the integral term I is , which is used to accumulate historical errors and eliminate the steady-state deviation caused by mechanical static friction; the differential term D is , used to predict the error change trend and suppress overshoot during rapid arm swing; S43 parameter tuning: Use the Ziegler-Nichols critical proportionality method to tune PID parameters.

5. The control method of a reconfigurable exoskeleton-assisted robot according to claim 4, characterized in that: Step S43 is as follows: gradually increase the proportional gain Until the system jitters, then based on the critical increase and oscillation period to calculate the final parameters , / , .

6. The control method of a reconfigurable exoskeleton-assisted robot according to claim 1, characterized in that: Step S5 is specifically as follows: Based on the fused Euler angle data and three-axis acceleration values, combined with the preset motor rotation range threshold, the weights of the three angles are adaptively allocated to identify the current upper and lower limb states of the human body, thereby switching the upper limb following mode, upper limb lifting mode, lower limb walking mode and human body carrying mode.

7. A reconfigurable exoskeleton-assisted robot, characterized in that: The control method according to any one of claims 1 to 6 is used for implementation, comprising: a backpack backboard (108) provided with a quick-release locking seat; a battery module (104) and a main control module (105) installed on the front of the backpack backboard (108); a detachable joint module connected to the backpack backboard via a quick-release locking seat; sensor modules distributed on the detachable joint module and the backpack backboard (108); a voice module (202) installed on the back of the backpack backboard (108) for receiving voice commands; wherein the main control module (105) controls the movement of the joint module in real time through the sensor module.

8. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium, characterized in that A computer program is stored, and when the program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images