Motion capture method and device of rehabilitation robot

Abstract

The invention provides a motion capture method and equipment of a rehabilitation robot. An IMU sensor in a passive glove can obtain sensor information such as a gyroscope, an accelerometer and a magnetometer in real time. The IMU sensors are placed on the palm and the fingertips of the five fingers of the passive glove, sensor data of all positions can be collected in real time, and the data are transmitted to the MCU to be processed. The MCU estimates the spatial attitude of each sensor through an algorithm and calculates the actual bending angle of each finger. According to the method, the posture data of each finger can be compared with the target rehabilitation action, feedback control is formed, the bending angle of the finger can be accurately captured, errors caused by sensor position deviation are effectively avoided, and higher posture estimation precision is achieved. The miniature IMU sensor is used, a flexible potentiometer or an external optical device is not needed, the environment interference resistance is high, use in a mobile scene is facilitated, and the requirements of the rehabilitation robot for comfort, portability and low cost are met.

Description

Technical Field

[0001] The invention relates to a motion capture method and equipment for a rehabilitation robot. Background Art

[0002] Traditional hand motion capture solutions usually use external vision systems, such as camera optical tracking and other methods. Although they are highly accurate, they are costly, have high requirements for ambient light, and are easily affected by occlusion of hand movements, making them inconvenient for rehabilitation applications.

[0003] Measuring the finger bending angle through a flexible potentiometer is also one of the traditional hand motion capture solutions, but it has problems such as uncomfortable wearing and accuracy affected by sensor fatigue.

[0004] Therefore, current technologies have deficiencies in accuracy, portability, anti-interference and cost when it comes to achieving high-precision hand motion capture, making it difficult to meet the needs of soft rehabilitation robot applications. Summary of the invention

[0005] The purpose of the present invention is to provide a motion capture method and device for a rehabilitation robot.

[0006] In order to solve the above problems, the present invention provides a motion capture method for a rehabilitation robot, comprising:

[0007] IMU sensors are installed on the palm and fingertips of each finger of the passive glove of the rehabilitation robot;

[0008] Collect data from each IMU sensor in real time and transmit it to the microcontroller unit;

[0009] The microcontroller unit estimates the spatial posture of each IMU sensor based on the data of each IMU sensor, and calculates the actual bending angle of each finger based on the spatial posture of each IMU sensor;

[0010] The actual bending angle of each finger of the passive glove is compared with the target bending angle of the target rehabilitation action to obtain an angle deviation value, and the actual bending angle of each finger of the passive glove is corrected based on the angle deviation value.

[0011] Furthermore, in the above method, the microcontroller unit estimates the spatial posture of each IMU sensor based on the data of each IMU sensor, and calculates the actual bending angle of each finger based on the spatial posture of each IMU sensor, including:

[0012] Real-time acquisition of raw sensor data from the passive glove through each IMU sensor;

[0013] The raw sensor data from a total of 6 IMU sensors on the palm and each fingertip are transmitted to the microcontroller unit via wires;

[0014] The microcontroller unit MCU performs preprocessing of calibration and filtering on the raw data of the sensor to obtain preprocessed sensor data;

[0015] Use the preprocessed sensor data to perform spatial attitude estimation to obtain a spatial attitude estimation result;

[0016] The three sensor data of each IMU sensor, namely angle change data, attitude data and angle, are fused by using a filtering algorithm to obtain the relative rotation quaternion q of each IMU sensor to describe the spatial attitude of each IMU sensor.

[0017] Based on the IMU sensor of the palm, calculate the relative rotation quaternion q of the IMU sensor of each finger tip The relative rotation quaternion q of the IMU sensor of the palm base The corresponding relative rotation quaternion difference q between rel =q base -1 qtip ;

[0018] The relative rotation quaternion difference q corresponding to each finger rel Convert to Euler angles to get the actual bending angle of each finger.

[0019] Furthermore, in the above method, the raw sensor data of each IMU sensor includes: three-axis accelerometer data (ax, ay, az), three-axis plus gyroscope data (gx, gy, gz) and three-axis magnetometer data (mx, my, mz).

[0020] Furthermore, in the above method, the microcontroller unit MCU calibrates the raw data of the sensor, including:

[0021] The three-axis accelerometer data, the three-axis plus gyroscope data and the three-axis magnetometer data are corrected for zero bias and scale factor respectively.

[0022] Furthermore, in the above method, using the preprocessed sensor data to perform spatial attitude estimation to obtain a spatial attitude estimation result includes:

[0023] The angle change data of each IMU sensor is calculated by integrating the gyroscope angular velocity over time in the preprocessed sensor data;

[0024] Using the accelerometer data in the preprocessed sensor data, calculate the current sensor's attitude data relative to the gravity direction;

[0025] The magnetometer data in the preprocessed sensor data is used to calculate the angle with the geomagnetic direction, and the angle is used to correct the drift caused by the gyroscope error integration.

[0026] Furthermore, in the above method, the filtering algorithm includes: Madgwick or Kalman filter.

[0027] Furthermore, in the above method, the actual bending angle of each finger of the passive glove is compared with the target bending angle of the target rehabilitation action.

[0028] The target bending angle of the standard rehabilitation action is collected through the active glove of the rehabilitation robot.

[0029] Furthermore, in the above method, the actual bending angle of each finger of the passive glove is compared with the target bending angle of the target rehabilitation action to obtain an angle deviation value, and the actual bending angle of each finger of the passive glove is corrected based on the angle deviation value, including:

[0030] In the pneumatically controlled passive glove, each finger air cavity is provided, each finger air cavity is connected to an air pump, and a solenoid valve is provided at each air path node between the air pump and each finger air cavity; a micro control unit is connected to the air pump and the solenoid valve respectively; and an air pressure sensor is installed in the connecting air path between the solenoid valve and the finger air cavity;

[0031] The microcontroller unit obtains the target air pressure value of the passive glove according to the air pressure sensor data of the passive glove and the following formula:

[0032] P(t)=k adaptive (t)·θ(t-β)+b+α·(P(t-Δt)-P feedback (t-Δt)),

[0033] Where t represents the current time, t-Δt represents the previous time; P(t) represents the target pressure value at the current time; k adaptive (t) is the adaptive proportional coefficient adjusted based on the pressure sensor feedback; β is the hysteresis compensation factor, θ(t+β) is the finger angle to be executed, where θ(t+β) represents the finger angle to be executed after the hysteresis compensation coefficient is introduced; b is the initial pressure offset; P(t-Δt) represents the target pressure value at the previous moment t-Δt, P feedback (t-Δt) is the actual feedback pressure value at the previous moment t-Δt; α is the feedback correction factor, which is used to calculate the pressure according to the deviation, that is, P(t-Δt)-P feedback (t-Δt) to adjust the air pressure target, the deviation refers to P(t-Δt)-P feedback (t-Δt);

[0034] The microcontroller unit detects the actual bending angle of each finger air cavity of the passive glove in real time through the IMU sensor carried in the passive glove. If the error value between the actual bending angle and the target bending angle of the target rehabilitation action is greater than a preset threshold, the microcontroller unit dynamically corrects the target air pressure value of each finger air cavity of the passive glove according to the difference between the actual bending angle and the target bending angle of the target rehabilitation action.

[0035] The microcontrol unit converts the corrected target air pressure value of each finger air cavity into an adjustment instruction of the air pump and the solenoid valve, and controls the opening or closing of the air pump and the solenoid valve based on the adjustment instruction.

[0036] According to another aspect of the present invention, there is further provided a computer-readable storage medium having computer-executable instructions stored thereon, wherein when the computer-executable instructions are executed by a processor, the processor is caused to: execute any of the methods described above.

[0037] According to another aspect of the present invention, there is also provided a computer device, comprising:

[0038] Processor; and

[0039] A memory arranged to store computer executable instructions, which when executed cause the processor to: perform any of the methods described above.

[0040] Compared with the prior art, the present invention has the following advantages:

[0041] Realize feedback control: The IMU sensor in the passive glove of the present invention can obtain sensor information such as gyroscopes, accelerometers and magnetometers in real time. The IMU sensors are placed on the palm of the passive glove and the fingertips of the five fingers, and can collect sensor data at various positions in real time and transmit these data to the MCU for processing. The MCU estimates the spatial posture of each sensor through an algorithm and calculates the actual bending angle of each finger. The posture data of each finger can be compared with the target rehabilitation action to form feedback control.

[0042] High-precision motion capture: By installing IMU sensors on the palm and fingertips of the passive glove, this system fuses multi-IMU sensor data and accurately captures the bending angle of the fingers through the correlation calculation between sensor position and joint structure, effectively avoiding the error caused by sensor position offset and achieving higher posture estimation accuracy.

[0043] Portable, comfortable and lightweight design: Using miniaturized IMU sensors, no flexible potentiometers or external optical devices are required, strong resistance to environmental interference, easy to use in mobile scenarios, and suitable for rehabilitation robots for comfort, portability and low cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flow chart of a motion capture method for a rehabilitation robot according to an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0046] In a typical configuration of the present application, the terminal, the device of the service network and the trusted party all include one or more processors (CPU), input / output interfaces, network interfaces and memories.

[0047] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0048] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include non-transitory media such as modulated data signals and carrier waves.

[0049] like Figure 1 As shown, the present invention provides a motion capture method for a rehabilitation robot, wherein the rehabilitation robot includes a passive glove and an active glove, including:

[0050] Step S1, installing IMU sensors on the palm and each fingertip of the passive glove of the rehabilitation robot;

[0051] Here, in the present invention, the active glove of the rehabilitation robot can be used to collect the target finger posture demonstrated by the user, and the passive glove is responsible for executing the action corresponding to the demonstrated target finger posture;

[0052] Step S2, collecting IMU sensor data at each position in real time and transmitting it to the microcontroller unit MCU;

[0053] Step S3, the microcontroller unit MCU estimates the spatial posture of each IMU sensor based on the IMU sensor data, and calculates the actual bending angle of each finger based on the spatial posture of each IMU sensor;

[0054] Preferably, step S3 may include:

[0055] Step S31, real-time acquisition of raw sensor data of the passive glove through each IMU sensor, the raw sensor data of each IMU sensor includes: three-axis accelerometer data (ax, ay, az), three-axis plus gyroscope data (gx, gy, gz) and three-axis magnetometer data (mx, my, mz).

[0056] Here, the IMU sensor can be a nine-axis IMU sensor; the nine-axis IMU sensor is an inertial measurement unit that integrates a three-axis gyroscope, a three-axis accelerometer, and a three-axis magnetometer; the nine-axis IMU sensor fuses the data of the above three sensors to more accurately and comprehensively estimate the motion state and posture of the object, including three-dimensional linear acceleration, three-dimensional angular velocity, three-dimensional rotation quaternion, vertical direction, and heading information. Compared with the common six-axis IMU, the nine-axis IMU adds a magnetometer, providing richer information.

[0057] Step S32, the raw sensor data of the six IMU sensors of the palm and each fingertip are transmitted to the microcontroller unit MCU via a wired manner;

[0058] In step S33, the microcontroller unit MCU performs preprocessing of calibration and filtering on the raw data of the sensor to obtain preprocessed sensor data.

[0059] Preferably, the micro control unit MCU calibrates the original data of the sensor, including: performing zero bias and scale factor correction on the three-axis accelerometer data, the three-axis plus gyroscope data and the three-axis magnetometer data, respectively.

[0060] Step S34, using the preprocessed sensor data to perform spatial attitude estimation to obtain a spatial attitude estimation result;

[0061] Preferably, step S34 may include: using the time integration of the gyroscope angular velocity in the preprocessed sensor data to calculate the angle change data of each IMU sensor; using the accelerometer data in the preprocessed sensor data to calculate the posture data of the current sensor relative to the gravity direction; using the magnetometer data in the preprocessed sensor data to calculate the angle with the geomagnetic direction, and the angle is used to correct the drift caused by the gyroscope error integration.

[0062] Step S35, using a filtering algorithm to fuse the three sensor data of each IMU sensor, namely, angle change data, posture data and angle, to obtain the relative rotation quaternion q of each IMU sensor to describe the spatial posture of each IMU sensor; preferably, the filtering algorithm may include: Madgwick or Kalman filter.

[0063] Step S36, using the IMU sensor of the palm as a reference, calculate the relative rotation quaternion q of the IMU sensor of each finger tip The relative rotation quaternion q of the IMU sensor of the palm base The corresponding relative rotation quaternion difference q between rel =q base -1 qtip ;

[0064] Here, each IMU sensor data can calculate a relative rotation quaternion q, which is used to represent the spatial posture of the IMU sensor, and then the relative rotation quaternion difference qrel = qbase-1qtip is obtained by comparing the fingertips and palms. Finally, this relative rotation quaternion qrel is converted into Euler angles to obtain the finger bending angle;

[0065] Step S37: The relative rotation quaternion difference q corresponding to each finger is rel Convert to Euler angle to get the actual bending angle of each finger;

[0066] Step S4, comparing the actual bending angle of each finger of the passive glove with the target bending angle of the target rehabilitation action to obtain an angle deviation value, and correcting the actual bending angle of each finger of the passive glove based on the angle deviation value.

[0067] Preferably, the target bending angle of the target rehabilitation action can be collected by the active glove of the rehabilitation robot, wherein the active glove can be used to collect the finger posture demonstrated by the user, and the passive glove is responsible for executing the action corresponding to the demonstrated finger posture;

[0068] Here, if this is used for feedback control, such as in a pneumatically controlled passive glove, it can be used to compare the actual bending angle of each finger of the passive glove with the target bending angle of the target rehabilitation action, obtain the angle deviation value, and realize closed-loop control. Due to the influence of factors such as software structure and airflow delay, the actual angle may deviate from the target angle. The angle information provided by the IMU sensor can be used as control accuracy compensation near the target angle. For example, if the target angle is 120° bending and the actual angle is 130°, the system can correct the error by adjusting the target angle.

[0069] Specifically, step S4 may include:

[0070] Step S41: In the pneumatically controlled passive glove, each finger air cavity can be set, each finger air cavity is connected to an air pump, and a solenoid valve is set at each air path node between the air pump and each finger air cavity; the micro control unit is connected to the air pump and the solenoid valve respectively; an air pressure sensor is installed in the connecting air path between the solenoid valve and the finger air cavity; and an IMU sensor is installed on the palm and each fingertip of the passive glove. Preferably, each finger air cavity can be set in 5 finger sleeves of a glove body, each finger sleeve is provided with a skeleton and a finger air cavity, and the skeleton is set outside the finger air cavity and is used to limit the radial expansion of the finger air cavity;

[0071] Here, the microcontroller unit can control the switch of the air pump and each solenoid valve to control the overall air system. Air pump: provides a stable air source for the system. Solenoid valve: an air node located between the air pump and each finger air cavity, which can control the on and off of the air flow by controlling the switch of the solenoid valve to ensure independent control of each finger air cavity. Finger air cavity: the air cavity structure of each finger can be deformed under different air pressures to achieve bending and stretching, so movement can be achieved through changes in air pressure.

[0072] Step S42, the micro control unit obtains the target air pressure value of the passive glove according to the air pressure sensor data of the passive glove;

[0073] Preferably, the microcontroller unit obtains the target air pressure value of the passive glove according to the air pressure sensor data of the passive glove through the following formula:

[0074] P(t)=k adaptive (t)·θ(t-β)+b+α·(P(t-Δt)-P feedback (t-Δt)),

[0075] Among them, t represents the current moment, and t-Δt represents the previous moment;

[0076] P(t) represents the target air pressure value at the current moment;

[0077] k adaptive (t) is an adaptive proportional coefficient adjusted based on the feedback of the air pressure sensor; preferably, an air pressure sensor is installed in the connecting air path between the solenoid valve and the finger air cavity; preferably, k adaptive (t) is dynamically updated in every Δt cycle, and each Δt is a cycle. In the real-time control process, based on the feedback pressure P feedback The difference between (t-Δt) and the target pressure P(t) is k adaptive (t) Dynamic adjustment; wherein, an incremental adjustment algorithm is used, when the feedback pressure P feedback When (t-Δt) is less than the target pressure P(t), gradually increase k adaptive(t) to increase the air pressure; otherwise, k is gradually reduced adaptive (t). Here, k adaptive (t) is the adaptive proportional coefficient adjusted based on the pressure sensor feedback, so that the target pressure gradually approaches the optimal pressure range that meets the patient's needs.

[0078] β is the hysteresis compensation factor, θ(t+β) is the finger angle to be executed, where θ(t+β) represents the finger angle to be executed after the hysteresis compensation coefficient is introduced; the finger angle θ(t+β) to be executed is offset by β time from the finger angle θ(t) in the active glove at the current time t, that is, the finger angle θ(t+β) to be executed obtained by the microcontroller unit is the finger angle to be executed β time before the current time t; preferably, β can be determined by experiment, the target air pressure value P(t) is set, and each finger air cavity starts to be inflated according to the target air pressure value P(t), and the timing is started after the inflation is completed, and the timing is stopped after the finger air cavity stops deforming, and the difference time between the start timing and the stop timing is determined as the hysteresis compensation factor β. Here, the present application introduces an advance compensation θ(t-β) to the finger bending angle θ(t) in the active glove, so that the system considers the influence of the hysteresis of the air cavity filling and discharging in advance, and reduces the hysteresis error caused by the deformation of the air cavity. β is the hysteresis compensation coefficient, and its value is adjusted based on the material properties of the air cavity and the response delay. Through this advance amount β, the microcontroller unit can fine-tune the air pressure before the attitude changes, reducing the error caused by lag.

[0079] b is the initial air pressure offset. Preferably, in the experimental stage, the target air pressure value P(t) of the air cavity under different bending angles is measured, and the target air pressure value P(t) is imported into MATLAB for curve fitting to obtain a suitable initial k adaptive (t) value and b value.

[0080] P(t-Δt) represents the target pressure value at the previous time t-Δt, P feedback (t-Δt) is the actual feedback pressure value at the previous moment t-Δt;

[0081] α is the feedback correction factor, which is used to adjust the deviation according to P(t-Δt)-P feedback (t-Δt) to adjust the air pressure target, the deviation refers to P(t-Δt)-P feedback (t-Δt); Preferably, in the experimental stage, the actual feedback pressure value P under the delayed condition of the air cavity filling and deflation under different target pressure values ​​P(t-Δt) can be measured. feedback (t-Δt), the deviation data is P(t-Δt)-P feedback (t-Δt), input the gradient descent optimization algorithm to adjust the α value to minimize the actual feedback pressure value P feedbackPreferably, the actual feedback pressure value P at the last moment t-Δt can be obtained by installing an air pressure sensor in the air path connecting the solenoid valve and the finger air cavity. feedback (t-Δt). Here, the correction term α·(P(t-Δt)-P feedback (t-Δt)), which compares the target pressure value P(t-Δt) at the previous moment with the actual feedback pressure value P feedback The deviation of (t-Δt) is used to adjust the target pressure value P(t) at the current moment in time. α is the feedback gain coefficient, which controls the response degree of the correction. The larger the α value, the faster the feedback correction, and the system can reach the target pressure more quickly. α is not the larger the better, and a suitable value should be taken, which can be determined through experiments.

[0082] Step S43, the microcontroller unit detects the actual bending angle of each finger air cavity of the passive glove in real time through the IMU sensor carried in the passive glove. If the error value between the actual bending angle and the target bending angle of the target rehabilitation action is greater than a preset threshold, the microcontroller unit dynamically adjusts the target air pressure value of the passive glove according to the difference between the actual bending angle and the target bending angle of the target rehabilitation action.

[0083] In step S43, the microcontroller unit converts the corrected target air pressure value of each finger air cavity into an adjustment instruction of the air pump and the solenoid valve, and controls the opening or closing of the air pump and the solenoid valve based on the adjustment instruction.

[0084] According to another aspect of the present invention, there is further provided a computer-readable storage medium having computer-executable instructions stored thereon, wherein when the computer-executable instructions are executed by a processor, the processor is caused to: execute any of the methods described above.

[0085] According to another aspect of the present invention, there is also provided a computer device, comprising:

[0086] Processor; and

[0087] A memory arranged to store computer executable instructions, which when executed cause the processor to: perform any of the methods described above.

[0088] In summary, the present invention has the following advantages:

[0089] Realize feedback control: IMU sensors can obtain information from sensors such as gyroscopes, accelerometers, and magnetometers in real time. IMU sensors are placed on the palm and fingertips of the five fingers to collect sensor data at various positions in real time and transmit the data to the MCU for processing. The MCU estimates the spatial posture of each sensor through an algorithm and calculates the actual bending angle of each finger. The posture data of each finger can be compared with the target rehabilitation action to form feedback control.

[0090] High-precision motion capture: By installing IMU sensors on the palm and fingertips of the passive glove, this system fuses multi-IMU sensor data and accurately captures the bending angle of the fingers through the correlation calculation between sensor position and joint structure, effectively avoiding the error caused by sensor position offset and achieving higher posture estimation accuracy.

[0091] Portable, comfortable and lightweight design: Using miniaturized sensors, no flexible potentiometers or external optical devices are required, strong resistance to environmental interference, easy to use in mobile scenarios, and suitable for rehabilitation robots for comfort, portability and low cost.

[0092] The detailed contents of the device embodiments of the present invention can be found in the corresponding parts of the method embodiments, which will not be described in detail here.

[0093] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

[0094] It should be noted that the present invention can be implemented in software and / or a combination of software and hardware, for example, can be implemented using an application specific integrated circuit (ASIC), a general purpose computer or any other similar hardware device. In one embodiment, the software program of the present invention can be executed by a processor to implement the steps or functions described above. Similarly, the software program of the present invention (including related data structures) can be stored in a computer readable recording medium, for example, a RAM memory, a magnetic or optical drive or a floppy disk and the like. In addition, some steps or functions of the present invention can be implemented using hardware, for example, as a circuit that cooperates with a processor to perform each step or function.

[0095] In addition, part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. The program instruction for calling the method of the present invention may be stored in a fixed or removable recording medium, and / or transmitted through a data stream in a broadcast or other signal-bearing medium, and / or

[0096] The computer program instructions are stored in a working memory of a computer device that runs according to the program instructions. Here, according to one embodiment of the present invention, a device is included, which includes a memory for storing computer program instructions and a processor for executing program instructions, wherein when the computer program instructions are executed by the processor, the device is triggered to run the methods and / or technical solutions based on the aforementioned multiple embodiments of the present invention.

[0097] It is obvious to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential features of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-restrictive, and the scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any figure mark in the claims should not be regarded as limiting the claims involved. In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the device claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

Claims

1. A motion capture method for a rehabilitation robot, characterized in that: include: IMU sensors are installed on the palm and fingertips of each finger of the passive glove of the rehabilitation robot; Collect data from each IMU sensor in real time and transmit it to the microcontroller unit; The microcontroller unit estimates the spatial posture of each IMU sensor based on the data of each IMU sensor, and calculates the actual bending angle of each finger based on the spatial posture of each IMU sensor; The actual bending angle of each finger of the passive glove is compared with the target bending angle of the target rehabilitation action to obtain an angle deviation value, and the actual bending angle of each finger of the passive glove is corrected based on the angle deviation value.

2. The motion capture method of a rehabilitation robot according to claim 1, characterized in that: The microcontroller unit estimates the spatial posture of each IMU sensor based on the data of each IMU sensor, and calculates the actual bending angle of each finger based on the spatial posture of each IMU sensor, including: Real-time acquisition of raw sensor data from the passive glove through each IMU sensor; The raw sensor data from a total of 6 IMU sensors on the palm and each fingertip are transmitted to the microcontroller unit via wires; The microcontroller unit MCU performs preprocessing of calibration and filtering on the raw data of the sensor to obtain preprocessed sensor data; Use the preprocessed sensor data to perform spatial attitude estimation to obtain a spatial attitude estimation result; The three sensor data of each IMU sensor, namely angle change data, attitude data and angle, are fused by using a filtering algorithm to obtain the relative rotation quaternion q of each IMU sensor to describe the spatial attitude of each IMU sensor. Based on the IMU sensor of the palm, calculate the relative rotation quaternion q of the IMU sensor of each finger tip The relative rotation quaternion q of the IMU sensor of the palm base The corresponding relative rotation quaternion difference q between rel =q base -1 qtip ; The relative rotation quaternion difference q corresponding to each finger rel Convert to Euler angles to get the actual bending angle of each finger.

3. The motion capture method of a rehabilitation robot as claimed in claim 2, characterized in that: The raw sensor data of each IMU sensor includes: three-axis accelerometer data (ax, ay, az), three-axis plus gyroscope data (gx, gy, gz), and three-axis magnetometer data (mx, my, mz).

4. The motion capture method of a rehabilitation robot as claimed in claim 2, characterized in that: The microcontroller unit MCU calibrates the raw data of the sensor, including: The three-axis accelerometer data, the three-axis plus gyroscope data and the three-axis magnetometer data are corrected for zero bias and scale factor respectively.

5. The motion capture method of a rehabilitation robot according to claim 1, characterized in that: Use the preprocessed sensor data to perform spatial attitude estimation to obtain spatial attitude estimation results, including: The angle change data of each IMU sensor is calculated by integrating the gyroscope angular velocity over time in the preprocessed sensor data; Using the accelerometer data in the preprocessed sensor data, calculate the current sensor's attitude data relative to the gravity direction; The magnetometer data in the preprocessed sensor data is used to calculate the angle with the geomagnetic direction, and the angle is used to correct the drift caused by the gyroscope error integration.

6. The motion capture method of a rehabilitation robot as claimed in claim 2, characterized in that: The filtering algorithm includes: Madgwick or Kalman filter.

7. The motion capture method of a rehabilitation robot according to claim 1, characterized in that: The actual bending angle of each finger of the passive glove is compared with the target bending angle of the target rehabilitation movement. The target bending angle of the standard rehabilitation action is collected through the active glove of the rehabilitation robot.

8. The motion capture method of a rehabilitation robot as claimed in claim 1, characterized in that: The actual bending angle of each finger of the passive glove is compared with the target bending angle of the target rehabilitation action to obtain an angle deviation value, and the actual bending angle of each finger of the passive glove is corrected based on the angle deviation value, including: In the pneumatically controlled passive glove, each finger air cavity is provided, each finger air cavity is connected to an air pump, and a solenoid valve is provided at each air path node between the air pump and each finger air cavity; a micro control unit is connected to the air pump and the solenoid valve respectively; and an air pressure sensor is installed in the connecting air path between the solenoid valve and the finger air cavity; The microcontroller unit obtains the target air pressure value of the passive glove according to the air pressure sensor data of the passive glove and the following formula: P(t)=k adaptive (t)·θ(t-β)+b+α·(P(t-Δt)-P feedback (t-Δt)), Where t represents the current time, t-Δt represents the previous time; P(t) represents the target pressure value at the current time; k adaptive (t) is the adaptive proportional coefficient adjusted based on the pressure sensor feedback; β is the hysteresis compensation factor, θ(t+β) is the finger angle to be executed, where θ(t+β) represents the finger angle to be executed after the hysteresis compensation coefficient is introduced; b is the initial pressure offset; P(t-Δt) represents the target pressure value at the previous moment t-Δt, P feedback (t-Δt) is the actual feedback pressure value at the previous moment t-Δt; α is the feedback correction factor, which is used to calculate the pressure according to the deviation, that is, P(t-Δt)-P feedback (t-Δt) to adjust the air pressure target, the deviation refers to P(t-Δt)-P feedback (t-Δt); The microcontroller unit detects the actual bending angle of each finger air cavity of the passive glove in real time through the IMU sensor carried in the passive glove. If the error value between the actual bending angle and the target bending angle of the target rehabilitation action is greater than a preset threshold, the microcontroller unit dynamically corrects the target air pressure value of each finger air cavity of the passive glove according to the difference between the actual bending angle and the target bending angle of the target rehabilitation action. The microcontrol unit converts the corrected target air pressure value of each finger air cavity into an adjustment instruction of the air pump and the solenoid valve, and controls the opening or closing of the air pump and the solenoid valve based on the adjustment instruction.

9. A computer-readable storage medium having computer-executable instructions stored thereon, wherein: When the computer executable instructions are executed by a processor, the processor is caused to: perform the method according to any one of claims 1 to 8.

10. A computer device, wherein: include: processor; as well as A memory arranged to store computer executable instructions which, when executed, cause the processor to: perform a method as claimed in any one of claims 1 to 8.

Images