Lower limb rehabilitation walking assistance system
By using PID control and multimodal data fusion in the lower limb rehabilitation walking assistance system, precise weight loss and safety support for the target subjects are achieved, solving the problem of inaccurate weight loss in existing technologies and improving the effectiveness of rehabilitation training.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- SHENZHEN INST OF ADVANCED TECH CHINESE ACAD OF SCI
- Filing Date
- 2024-12-10
- Publication Date
- 2026-05-15
AI Technical Summary
Current lower limb rehabilitation training cannot achieve precise weight loss, especially for patients with grade 2-4 muscle strength, and cannot effectively assist them in strengthening muscle strength and improving exercise quality during the recovery process.
The lower limb rehabilitation walking assistance system uses a movable device, an adjustable device, and a human fixation device, combined with a torque motor and a force feedback motor. It uses PID control to adjust the opening and closing angle and the supporting force to achieve precise weight reduction. The target weight reduction is calculated by a host computer to reduce the computational burden on the system.
It enables precise weight reduction during lower limb rehabilitation training, improves the safety and comfort of training, ensures that the target subject receives appropriate support under different movement states, and improves training effectiveness.
Smart Images
Figure CN2024137994_15052026_PF_FP_ABST
Abstract
Description
Lower limb rehabilitation walking aid system
[0001] This application claims priority to Chinese Patent Application No. 202411572570.8, filed with the Chinese Patent Office on November 6, 2024, the entire contents of which are incorporated herein by reference. Technical Field
[0002] This application relates to the field of physiotherapy technology, such as a lower limb rehabilitation assistive system. Background Technology
[0003] Stroke or cerebrovascular accident patients, patients recovering from trauma, patients with neuromuscular diseases, and the elderly need rehabilitation training to improve their motor abilities due to changes in muscle strength or trauma.
[0004] According to muscle strength grading, the Medical Research Council (MRC) scoring system is typically used, which divides muscle strength into grades 0 to 5. Grade 0 indicates no muscle contraction at all; Grade 1 indicates slight muscle contraction but no joint movement; Grade 2 indicates joint movement that does not counteract gravity; Grade 3 indicates joint movement that counteracts gravity but cannot resist external forces; Grade 4 indicates joint movement that can resist some external forces; and Grade 5 indicates normal muscle strength. Grades 0 to 3 have weak mobility and poor resistance to external forces, therefore requiring external assistance for movement.
[0005] Specifically, for lower limb rehabilitation training (which requires voluntary joint movement), grade 2 patients are the primary target group for weight-loss techniques. This is because these patients are unable to stand independently or resist gravity, and weight loss can help them restart exercise and rehabilitation training. Patients with grade 3 to 4 muscle strength can resist gravity, but are still limited during the recovery process. Weight loss can help them strengthen their muscles and improve the quality of movement.
[0006] However, precise weight loss is currently not possible during lower limb rehabilitation training, which urgently needs to be addressed. Summary of the Invention
[0007] This application provides a lower limb rehabilitation walking aid system, which may include: a lower limb rehabilitation walking aid; wherein...
[0008] The lower limb rehabilitation walking aid includes a movable device that can move on the ground, a first adjustment device, a second adjustment device, a human body fixation device, and a main control device. One end of the movable device is V-shaped and rotatably connected to the first adjustment device, and the other end is V-shaped and rotatably connected to the second adjustment device. The first adjustment device and the second adjustment device are X-shaped and rotatably connected. The first adjustment device is equipped with a first torque motor and a force feedback motor. The force feedback motor is rigidly connected to the human body fixation device. The second adjustment device is equipped with a second torque motor.
[0009] The main control unit includes a pressure acquisition unit, a target weight reduction acquisition unit, a motor output adjustment parameter acquisition unit, and a torque motor output adjustment unit; among which,
[0010] The pressure acquisition unit is configured to acquire the pressure applied to the human body fixation device by the target object undergoing lower limb rehabilitation training by wearing the human body fixation device through a force feedback motor.
[0011] The target weight reduction unit is set to obtain the target weight reduction of the target object;
[0012] The motor output adjustment parameter acquisition unit is set to perform proportional-integral-derivative control based on the error between pressure and target weight reduction to obtain the motor output adjustment parameters;
[0013] The torque motor output adjustment unit is configured to adjust the output of at least one of the first torque motor and the second torque motor according to the motor output adjustment parameters, so as to adjust the opening and closing angle between the first adjustment device and the second adjustment device. The opening and closing angle is negatively correlated with the lifting force provided to the target object through the human body fixation device, and the lifting force is used to resist the gravity of the target object. Attached Figure Description
[0014] Figure 1 is a structural block diagram of a lower limb rehabilitation assistive walking system provided in an embodiment of this application;
[0015] Figure 2 is a schematic diagram of a lower limb rehabilitation walking aid in a lower limb rehabilitation walking aid system provided in an embodiment of this application;
[0016] Figure 3 is a structural block diagram of a lower limb rehabilitation walking assistance system provided in an embodiment of this application;
[0017] Figure 4 is a structural block diagram of a lower limb rehabilitation assistive walking system provided in an embodiment of this application;
[0018] Figure 5A is a schematic diagram of the IMU sensor C and sEMG acquisition unit in a lower limb rehabilitation walking assistance system provided in an embodiment of this application;
[0019] Figure 5B is a schematic diagram of a plantar pressure measurement unit in a lower limb rehabilitation walking assistance system provided in an embodiment of this application;
[0020] Figure 6A is a schematic diagram of multimodal data acquisition in a lower limb rehabilitation walking assistance system provided in an embodiment of this application;
[0021] Figure 6B is a structural block diagram of an electromyography acquisition unit corresponding to Figure 6A in a lower limb rehabilitation assistive walking system provided in an embodiment of this application;
[0022] Figure 6C is a structural block diagram of an IMU acquisition unit corresponding to Figure 6A in a lower limb rehabilitation walking assistance system provided in an embodiment of this application;
[0023] Figure 6D is a structural block diagram of a plantar pressure measurement unit corresponding to Figure 6A in a lower limb rehabilitation walking assistance system provided in an embodiment of this application;
[0024] Figure 7 is a structural block diagram of a lower limb rehabilitation assistive walking system provided in an embodiment of this application;
[0025] Figure 8 is a flowchart of a lower limb rehabilitation assistive system provided in an embodiment of this application. Detailed Implementation
[0026] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. The same applies to "target," "original," etc., and will not be repeated here. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, for example, including, in addition to processes, methods, systems, products, or devices that include the series of steps or units shown in the embodiments of this application, other processes, methods, systems, products, and devices that are not explicitly listed in this series of steps or units, or other steps or units inherent to these processes, methods, systems, products, or devices.
[0027] Figure 1 is a structural block diagram of a lower limb rehabilitation assistive walking system provided in an embodiment of this application. This embodiment is applicable to situations involving weight loss during lower limb rehabilitation training.
[0028] Referring to Figure 1, the lower limb rehabilitation walking aid system according to an embodiment of this application includes: a lower limb rehabilitation walking aid A; wherein,
[0029] The lower limb rehabilitation walking aid A includes a movable device A1 that can move on the ground, a first adjustment device A2, a second adjustment device A3, a human body fixation device A4, and a main control device A5. One end of the movable device A1 is V-shapedly rotatably connected to the first adjustment device A2 and the other end is V-shapedly rotatably connected to the second adjustment device A3. The first adjustment device A2 and the second adjustment device A3 are X-shapedly rotatably connected. The first adjustment device A2 is equipped with a first torque motor and a force feedback motor. The force feedback motor is rigidly connected to the human body fixation device A4. The second adjustment device A3 is equipped with a second torque motor.
[0030] The main control unit A5 includes a pressure acquisition unit A51, a target weight reduction acquisition unit A52, a motor output adjustment parameter acquisition unit A53, and a torque motor output adjustment unit A54; among which,
[0031] The pressure acquisition unit A51 is configured to acquire the pressure applied to the human body fixation device A4 by the target object undergoing lower limb rehabilitation training by wearing the human body fixation device A4 through a force feedback motor.
[0032] The target weight reduction acquisition unit A52 is set to obtain the target weight reduction of the target object;
[0033] The motor output adjustment parameter acquisition unit A53 is set to perform proportional-integral-derivative control based on the error between pressure and target weight reduction to obtain the motor output adjustment parameters;
[0034] The torque motor output adjustment unit A54 is configured to adjust the output of at least one of the first torque motor and the second torque motor according to the motor output adjustment parameters, so as to adjust the opening and closing angle between the first adjustment device A2 and the second adjustment device A3, wherein the opening and closing angle is negatively correlated with the lifting force provided to the target object through the human body fixing device A4, and the lifting force is used to resist the gravity of the target object.
[0035] To better illustrate the lower limb rehabilitation walking aid A described above, Figure 2 is used as an example for explanation.
[0036] For example, from a hardware perspective, the lower limb rehabilitation walking aid A includes a movable device A1, a first adjustment device A2, a second adjustment device A3, a human body fixation device A4, two torque motors (i.e., the first torque motor described above) A6 mounted on the first adjustment device A2, two torque motors (i.e., the second torque motor described above) A6 mounted on the second adjustment device A3, and a force feedback motor A7. Based on this, the main control device A5 can be understood as the software device in the lower limb rehabilitation walking aid A, and can be implemented by a microcontroller unit (MCU) embedded in the lower limb rehabilitation walking aid A.
[0037] Based on this, the target object can be understood as an object wearing a human fixation device A4. The lower limb rehabilitation walking aid A is mainly responsible for providing auxiliary support to the target object, enabling the target object to walk, thereby carrying out lower limb rehabilitation training. Through the force feedback motor A7, the pressure applied by the target object to the human fixation device A4 can be obtained. Since there is an interaction force between this pressure and the supporting force actually provided to the target object through the human fixation device A4, the supporting force can be characterized by this pressure.
[0038] Target weight reduction can be understood as the expected reduction in gravity for the target individual during lower limb rehabilitation training, i.e., the expected support force provided to the target individual. This target weight reduction can be calculated by the lower limb rehabilitation walking aid A, or by other devices used in conjunction with the lower limb rehabilitation walking aid A (such as a host computer), which can be set according to actual needs and is not specifically limited here.
[0039] It should be noted that the lower limb rehabilitation walking aid A provides support to the target object through the torque motor A6. However, because there may be a difference between the actual output and the expected output of the torque motor A6, there may be a difference between the actual support provided by the lower limb rehabilitation walking aid A to the target object and the expected support (i.e. the target weight loss), which may result in the inability to achieve precise weight loss during rehabilitation training.
[0040] To address the aforementioned technical problems, this application embodiment employs Proportional-Integral-Derivative (PID) control based on the error between pressure and target weight reduction to obtain motor output adjustment parameters. These parameters can be understood as parameters used to adjust the output of torque motor A6, and further, as parameters used to adjust the output of force feedback motor A7. Based on this, exemplarily, the PID control can consist of the following three parts: Proportional (P) control: providing an output proportional to the error for rapid response to changes in the target object's state; Integral (I) control: eliminating accumulated errors by compensating for long-term deviations through integral processing of the error, ensuring long-term accuracy; Derivative (D) control: predictively adjusting the rate of change of the error to suppress oscillations, providing smoother weight reduction adjustment, and ensuring the stability of target object training.
[0041] Optionally, the output of at least one of the first torque motor A6 and the second torque motor A6 can be adjusted according to the motor output adjustment parameters to adjust the opening angle Angle between the first adjustment device A2 and the second adjustment device A3, thereby adjusting the opening height between the movable device A1 and the second adjustment device A3. It should be noted that a smaller opening angle Angle results in a higher opening height, thus providing a greater upward force to the target object through the human body restraint device A4, meaning the target object resists more gravity; conversely, a larger opening angle Angle results in a lower opening height, thus providing a smaller upward force to the target object through the human body restraint device A4, meaning the target object resists less gravity. In other words, the human body restraint device A4 serves to protect the target object and provide upward force to reduce weight when the opening angle Angle and opening height change; the opening angle Angle and the provided upward force are negatively correlated.
[0042] The lower limb rehabilitation walking aid described in this application uses PID control to adjust the output of the torque motor by measuring the error between the target weight loss and the actual supporting force provided to the target object, thereby ensuring the accuracy of the torque motor output and achieving precise weight loss during lower limb rehabilitation training.
[0043] In one optional technical solution, the main control device further includes a force feedback motor output adjustment unit: wherein,
[0044] The force feedback motor output adjustment unit is configured to adjust the output of the force feedback motor according to the motor output adjustment parameters in order to control the human body fixation device to rotate around the target object.
[0045] In this scenario, if the supporting force on the target object decreases, the target object may tilt forward; conversely, if the supporting force increases, the target object may tilt backward. Therefore, to ensure the stability of the target object's center of gravity, when the force feedback motor is rigidly connected to the human restraint device, the output of the force feedback motor can be adjusted according to the motor's output adjustment parameters. This allows the human restraint device to rotate around the target object, ensuring the target object remains upright. For example, if the supporting force decreases, the human restraint device is controlled to rotate behind the target object to pull it up; otherwise, it is controlled to rotate behind the target object to support it.
[0046] The above technical solution, on the one hand, can maintain the stability of the target object's center of gravity by adjusting the output of the force feedback motor, thereby ensuring the safety of the target object during the lower limb rehabilitation training process.
[0047] On the other hand, the coordinated operation between the torque motor and the force feedback motor ensures that the target subject receives proper support throughout the entire lower limb rehabilitation training process, effectively improving safety and comfort.
[0048] Figure 3 is a structural block diagram of a lower limb rehabilitation walking aid system provided in an embodiment of this application. This embodiment is an optimization based on the above-mentioned technical solutions. Optionally, in this embodiment, the system further includes: a host computer; wherein the host computer includes: a control command sending unit; wherein the control command sending unit is configured to send control commands to the lower limb rehabilitation walking aid; and the target weight loss obtaining unit is configured to analyze the received control commands in response to obtain the target weight loss of the target object. The explanations of terms that are the same as or corresponding to those in the above embodiments are not repeated here.
[0049] For example, referring to Figure 3, the lower limb rehabilitation walking assistance system described in this embodiment includes: a lower limb rehabilitation walking aid A and a host computer B; wherein,
[0050] The lower limb rehabilitation walking aid A includes a movable device A1 that can move on the ground, a first adjustment device A2, a second adjustment device A3, a human body fixation device A4, and a main control device A5. One end of the movable device A1 is V-shapedly rotatably connected to the first adjustment device A2 and the other end is V-shapedly rotatably connected to the second adjustment device A3. The first adjustment device A2 and the second adjustment device A3 are X-shapedly rotatably connected. The first adjustment device A2 is equipped with a first torque motor and a force feedback motor. The force feedback motor is rigidly connected to the human body fixation device A4. The second adjustment device A3 is equipped with a second torque motor.
[0051] The main control unit A5 includes a pressure acquisition unit A51, a target weight reduction acquisition unit A52, a motor output adjustment parameter acquisition unit A53, and a torque motor output adjustment unit A54;
[0052] The host computer B includes: a control command sending unit B1; wherein,
[0053] The control command sending unit B1 is configured to send control commands to the lower limb rehabilitation walking aid A;
[0054] The pressure acquisition unit A51 is configured to acquire the pressure applied to the human body fixation device A4 by the target object undergoing lower limb rehabilitation training by wearing the human body fixation device A4 through a force feedback motor.
[0055] The target weight reduction unit A52 is configured to respond to received control commands, analyze the control commands, and obtain the target weight reduction of the target object.
[0056] The motor output adjustment parameter acquisition unit A53 is set to perform proportional-integral-derivative control based on the error between pressure and target weight reduction to obtain the motor output adjustment parameters;
[0057] The torque motor output adjustment unit A54 is configured to adjust the output of at least one of the first torque motor and the second torque motor according to the motor output adjustment parameters, so as to adjust the opening and closing angle between the first adjustment device A2 and the second adjustment device A3, wherein the opening and closing angle is negatively correlated with the lifting force provided to the target object through the human body fixing device A4, and the lifting force is used to resist the gravity of the target object.
[0058] The technical solution of this application embodiment calculates the target weight loss through a host computer and generates corresponding control instructions. Then, the control instructions are sent to the lower limb rehabilitation walking aid so that the lower limb rehabilitation walking aid can analyze the target weight loss based on the control instructions. Compared with the lower limb rehabilitation walking aid calculating the target weight loss, this can reduce the calculation pressure on the lower limb rehabilitation walking aid, thereby allowing it to focus more on PID control and further achieve precise weight loss.
[0059] Figure 4 is a structural block diagram of a lower limb rehabilitation assistive walking system provided in an embodiment of this application. This embodiment is an optimization based on the above-described technical solutions. Optionally, in this embodiment, the host computer further includes: a data acquisition unit and a control command generation unit; wherein, the data acquisition unit is configured to acquire training data of the target object; the control command generation unit is configured to determine the motion state of the target object based on the training data, and generate control commands based on the motion state. Explanations of terms that are the same as or corresponding to those in the above embodiments will not be repeated here.
[0060] For example, referring to Figure 4, the lower limb rehabilitation walking assistance system described in this embodiment includes: a lower limb rehabilitation walking aid A and a host computer B; wherein,
[0061] The lower limb rehabilitation walking aid A includes a movable device A1 that can move on the ground, a first adjustment device A2, a second adjustment device A3, a human body fixation device A4, and a main control device A5. One end of the movable device A1 is V-shapedly rotatably connected to the first adjustment device A2 and the other end is V-shapedly rotatably connected to the second adjustment device A3. The first adjustment device A2 and the second adjustment device A3 are X-shapedly rotatably connected. The first adjustment device A2 is equipped with a first torque motor and a force feedback motor. The force feedback motor is rigidly connected to the human body fixation device A4. The second adjustment device A3 is equipped with a second torque motor.
[0062] The main control unit A5 includes a pressure acquisition unit A51, a target weight reduction acquisition unit A52, a motor output adjustment parameter acquisition unit A53, and a torque motor output adjustment unit A54;
[0063] The host computer B includes a data acquisition unit B3, a control command generation unit B2, and a control command sending unit B1; wherein,
[0064] Data acquisition unit B3 is configured to acquire training data of the target object;
[0065] The control command generation unit B2 is configured to determine the motion state of the target object based on the training data, and generate control commands based on the motion state.
[0066] The control command sending unit B1 is configured to send control commands to the lower limb rehabilitation walking aid A;
[0067] The pressure acquisition unit A51 is configured to acquire the pressure applied to the human body fixation device A4 by the target object undergoing lower limb rehabilitation training by wearing the human body fixation device A4 through a force feedback motor.
[0068] The target weight reduction unit A52 is configured to respond to received control commands, analyze the control commands, and obtain the target weight reduction of the target object.
[0069] The motor output adjustment parameter acquisition unit A53 is set to perform proportional-integral-derivative control based on the error between pressure and target weight reduction to obtain the motor output adjustment parameters;
[0070] The torque motor output adjustment unit A54 is configured to adjust the output of at least one of the first torque motor and the second torque motor according to the motor output adjustment parameters, so as to adjust the opening and closing angle between the first adjustment device A2 and the second adjustment device A3, wherein the opening and closing angle is negatively correlated with the lifting force provided to the target object through the human body fixing device A4, and the lifting force is used to resist the gravity of the target object.
[0071] The acquisition of training data for the target object can be understood as data related to the target object's lower limb rehabilitation training process. Based on this, and considering the application scenarios that may be involved in the embodiments of this application, the training data may optionally be multimodal data, which may include at least two of the following: lower limb surface electromyography signals, lower limb inertial measurement signals, plantar pressure signals, and actual body weight. Of course, the training data can also be unimodal data, which can be selected according to actual needs and is not specifically limited here.
[0072] Based on this, and considering the application scenarios that may be involved in the embodiments of this application, optionally, when the training data is multimodal data, the above system may further include at least one of: an inertial measurement unit (IMU) sensor (or IMU acquisition unit), a surface electromyography (sEMG) acquisition unit, and a plantar pressure measurement unit.
[0073] The data acquisition unit is configured to acquire multimodal data of the target object through at least two of the following steps:
[0074] The inertial measurement unit (IMU) sensor acquires the inertial measurement signals of the target object's lower limbs.
[0075] The surface electromyography (EMG) signal of the target object's lower limb is acquired through the surface EMG acquisition unit.
[0076] The plantar pressure signal of the target object is acquired through the plantar pressure measurement unit; and...
[0077] Obtain the actual weight of the target object. This actual weight can be measured directly using a scale and then input into the host computer.
[0078] For example, Figure 5A shows the IMU sensor C and sEMG acquisition unit D, Figure 5B shows the plantar pressure measurement unit, and Figures 6A-6D show the multimodal data acquisition process. For example,
[0079] (1) sEMG acquisition unit: See Figures 6A and 6B. The signal amplified by the operational amplifier is processed by the MCU and the corresponding signal is uploaded to the host computer via Bluetooth serial communication for further processing.
[0080] (2) IMU acquisition unit: Referring to Figures 6A and 6C, the lower limb joint angle changes and lower limb movement trajectories acquired by the IMU sensor are transmitted via serial communication and Bluetooth, and the corresponding signals are uploaded to the host computer for subsequent processing. It may also include filtering circuits and calibration circuits (such as calibrating temperature difference and zero drift).
[0081] (3) Plantar pressure measurement unit: Referring to Figures 6A and 6D, the plantar pressure signal is acquired by a pressure sensor and sampled by a multi-channel data acquisition circuit and an analog-to-digital converter (ADC) array. The signal is then transmitted via serial communication and Bluetooth and uploaded to the host computer for further processing. It may also include a filtering circuit and an operational amplifier circuit.
[0082] The power module may include a 3.3V voltage regulator circuit and a lithium battery charging and discharging circuit.
[0083] The signal amplification circuit can use a three-op-amp differential amplifier circuit to amplify the electrical signal.
[0084] The signal filtering circuit filters the acquired electrical signal based on its characteristics to obtain a high-quality signal.
[0085] The Bluetooth module can upload the corresponding signals to the host computer via transparent transmission, adopting a slave mode.
[0086] The electromyography (EMG) acquisition electrodes can be dry electrode pads. Each EMG acquisition unit requires a pair of dry electrode pads. The electrode pairs are worn on the rectus femoris, sartorius, vastus medialis, and vastus lateralis muscles of the target subject's left and right legs.
[0087] Multi-channel data acquisition circuits commonly use multiplexers (MUX) to switch between multiple sensors or use multi-channel ADCs to acquire data from multiple sensors at once.
[0088] In practical applications, optionally, the host computer can perform data preprocessing after acquiring multimodal data. For example, data preprocessing mainly includes:
[0089] Denoising techniques, such as filtering and wavelet transform, are employed. High-pass filters remove low-frequency noise, such as sensor drift or slow changes, making them suitable for extracting the dynamic components of lower limb IMU signals. Band-pass filters are used to process sEMG signals, extracting muscle activity features. Moving average filtering is used to smooth the signal, reducing random noise and making it suitable for smoothing pressure sensors. Wavelet transform is used to analyze non-stationary signals, such as sEMG and lower limb IMU signals. Wavelet transform can effectively separate noise from useful signals, making it suitable for processing complex signals with multiple frequency components.
[0090] Data cleaning can detect and process abnormal signals, remove or replace them, and process missing signals, including deletion and interpolation.
[0091] Optionally, the motion state of the target object is determined based on the training data, and then control commands are generated based on the motion state. In this embodiment, the motion state may optionally include at least one of gait stability, muscle fatigue, and center of gravity changes, in order to adjust the weight-reduction support of the lower limb rehabilitation walking aid.
[0092] The technical solution of this application embodiment can achieve adaptive weight reduction by determining the target weight reduction that matches the motion state, thereby ensuring that the target object can obtain appropriate support force in different motion states and avoiding excessive or insufficient weight reduction; in particular, the above-mentioned adaptive weight reduction process can be carried out in real time, thereby better ensuring that the target object is always under the optimal load and effectively improving the training effect.
[0093] An optional technical solution involves controlling a target object with a tendency to fall, where the motion indicates that the object is in a fall-prone state. The control commands include a weight-adjustment sub-command and a support force adjustment sub-command, and the main control device further includes a support force adjustment unit.
[0094] The target weight reduction unit is configured to respond to the received load adjustment sub-instruction, analyze the load adjustment sub-instruction, and obtain the target weight reduction of the target object.
[0095] The support force adjustment unit is configured to adjust the output of the force feedback motor in response to the received support force adjustment sub-command, so as to enable the human body fixation device to rotate around the target object.
[0096] The above-mentioned technical solution, when judging from the degree of muscle fatigue and gait stability that the target object has a tendency to fall, can provide additional support to the target object by adjusting the output of the force feedback motor, so as to ensure the safety of the target object during the lower limb rehabilitation training process and avoid accidental injury.
[0097] Another optional technical solution, the host computer, also includes a personalized training scheme acquisition unit; among which,
[0098] The personalized training scheme acquisition unit is set to acquire the personalized training scheme of the target object;
[0099] The control command generation unit is configured to determine the motion state of the target object based on the fusion results, and generate control commands based on the motion state and the personalized training plan.
[0100] The above-mentioned technical solution, by combining personalized training programs to determine the target weight loss, ensures the personalization of the lower limb rehabilitation training for the target individuals, thereby improving the training effect.
[0101] In addition, the host computer may optionally include: a rehabilitation training record acquisition unit and a personalized training plan generation unit;
[0102] The rehabilitation training record acquisition unit is set to acquire the rehabilitation training records of the target object during the rehabilitation training process using the lower limb rehabilitation walking aid;
[0103] The personalized training program generation unit is set to generate personalized training programs based on rehabilitation training records.
[0104] The above-mentioned technical solution can generate personalized training plans using rehabilitation training records. Based on this, the personalized training plans can be adjusted by combining multimodal data, thereby adapting to the different training progress and training needs of the target individuals, enabling them to obtain rehabilitation training that is most suitable for their own situation and improving the training effect.
[0105] Figure 7 is a structural block diagram of a lower limb rehabilitation assistive system provided in an embodiment of this application. This embodiment is an optimization based on the above-described technical solutions. In this embodiment, optionally, the training data is multimodal data, which includes at least two types of unimodal data; the control command generation unit includes a fusion subunit and a control command generation subunit; wherein, the fusion subunit is configured to fuse with at least two types of unimodal data to obtain a fusion result; the control command generation subunit is configured to determine the motion state of the target object based on the fusion result, and generate control commands based on the motion state. The explanations of terms that are the same as or corresponding to those in the above embodiments are not repeated here.
[0106] For example, referring to Figure 7, the lower limb rehabilitation walking assistance system described in this embodiment includes: a lower limb rehabilitation walking aid A and a host computer B; wherein,
[0107] The lower limb rehabilitation walking aid A includes a movable device A1 that can move on the ground, a first adjustment device A2, a second adjustment device A3, a human body fixation device A4, and a main control device A5. One end of the movable device A1 is V-shapedly rotatably connected to the first adjustment device A2 and the other end is V-shapedly rotatably connected to the second adjustment device A3. The first adjustment device A2 and the second adjustment device A3 are X-shapedly rotatably connected. The first adjustment device A2 is equipped with a first torque motor and a force feedback motor. The force feedback motor is rigidly connected to the human body fixation device A4. The second adjustment device A3 is equipped with a second torque motor.
[0108] The main control unit A5 includes a pressure acquisition unit A51, a target weight reduction acquisition unit A52, a motor output adjustment parameter acquisition unit A53, and a torque motor output adjustment unit A54;
[0109] The host computer B includes a data acquisition unit B3, a control command generation unit B2, and a control command sending unit B1. The control command generation unit B2 includes a fusion subunit B21 and a control command generation subunit B22; wherein,
[0110] Data acquisition unit B3 is configured to acquire multimodal data of the target object, wherein the multimodal data includes at least two types of unimodal data;
[0111] The fusion subunit B21 is configured to fuse with at least two single-modal data to obtain a fusion result;
[0112] The control command generation subunit B22 is configured to determine the motion state of the target object based on the fusion result, and generate control commands based on the motion state.
[0113] The control command sending unit B1 is configured to send control commands to the lower limb rehabilitation walking aid A;
[0114] The pressure acquisition unit A51 is configured to acquire the pressure applied to the human body fixation device A4 by the target object undergoing lower limb rehabilitation training by wearing the human body fixation device A4 through a force feedback motor.
[0115] The target weight reduction unit A52 is configured to respond to received control commands, analyze the control commands, and obtain the target weight reduction of the target object.
[0116] The motor output adjustment parameter acquisition unit A53 is set to perform proportional-integral-derivative control based on the error between pressure and target weight reduction to obtain the motor output adjustment parameters;
[0117] The torque motor output adjustment unit A54 is configured to adjust the output of at least one of the first torque motor and the second torque motor according to the motor output adjustment parameters, so as to adjust the opening and closing angle between the first adjustment device A2 and the second adjustment device A3, wherein the opening and closing angle is negatively correlated with the lifting force provided to the target object through the human body fixing device A4, and the lifting force is used to resist the gravity of the target object.
[0118] The process involves fusing data with at least two single-modal data to obtain a fusion result, and then determining the motion state based on the fusion result. This fusion result helps to more comprehensively evaluate the motion state.
[0119] Based on this, and considering the application scenarios that may be involved in the embodiments of this application, the optional fusion schemes include data-level fusion schemes, feature-level fusion schemes, and decision-level fusion schemes. Among them,
[0120] Data-level fusion schemes can be understood as directly fusing the raw signals acquired by the sensors. For example, features from plantar pressure signals, sEMG signals, and lower limb IMU signals are combined into a single feature vector for analysis. This fusion scheme retains all information but is susceptible to noise and signal redundancy. Feature-level fusion schemes first extract features from each sensor and then fuse these features. This fusion scheme reduces noise while maintaining signal diversity. For example, muscle activity features from sEMG, acceleration features from IMU, and gait features from plantar pressure can be combined to assess the motor abilities of a target. Decision-level fusion schemes involve each sensor analyzing and making decisions independently, then fusing the decisions from each sensor. This fusion scheme is suitable for distributed decision-making in multi-sensor systems.
[0121] Optional fusion solutions include machine learning algorithms and deep learning models. Machine learning algorithms, such as decision trees, support vector machines (SVM), and random forests, can perform classification or regression analysis based on multimodal features to determine the motion state and load requirements of the target object in real time. Deep learning models, such as convolutional neural networks (CNN) and recurrent neural networks (RNN), can automatically extract high-dimensional features from multimodal data to achieve more accurate state recognition and load adjustment.
[0122] Optional fusion schemes include Kalman filtering, which can be used to fuse lower limb IMU signals and plantar pressure signals to improve the accuracy of gait analysis and posture analysis; weighted fusion can also be used, which assigns different weights to the features of different sensors, such as giving higher weight to sEMG signals when detecting fatigue, and giving higher weight to lower limb IMU signals when detecting posture, etc., etc., without specific limitations.
[0123] The technical solution of this application embodiment, by collecting and fusing multimodal data, can provide a comprehensive assessment of the motion state, improve the accuracy of motion state assessment, and thus improve the accuracy of target weight loss determination.
[0124] One optional technical solution is that the fusion subunit includes a single-modal feature acquisition unit and a feature fusion unit; wherein,
[0125] The single-modal feature acquisition unit is configured to extract features from each of at least two types of single-modal data to obtain single-modal features;
[0126] The feature fusion unit is configured to fuse all single-modal features corresponding to at least two types of single-modal data to obtain the fusion result.
[0127] The feature-level fusion scheme described above can reduce noise while maintaining signal diversity.
[0128] Based on this, an optional single-modal feature acquisition unit is configured to perform feature extraction on the lower limb surface electromyography (EMG) signal when the single-modal data is the lower limb surface EMG signal, so as to obtain a single-modal feature that characterizes at least one of the activation degree and fatigue status of the lower limb muscles of the target object.
[0129] For example, sEMG signals are used to analyze the activation level and / or fatigue status of lower limb muscles. Feature extraction schemes may include at least one of time-domain features, frequency features, and time-frequency threshold features. For example,
[0130] Time-domain characteristics: (1) Mean and variance: used to analyze the overall level and fluctuation of muscle activity; (2) Root mean square (RMS): reflects the intensity of muscle contraction and is a commonly used characteristic for analyzing muscle workload; (3) Integral electromyography (IEMG): used to assess the overall intensity of muscle activity.
[0131] Frequency domain characteristics: (1) Intermediate frequency (MDF) and average power frequency (MPF): These two characteristics can reflect the fatigue state of muscles. As fatigue increases, the signal frequency usually decreases.
[0132] Time-frequency domain characteristics: (1) Wavelet transform and short-time Fourier transform: The frequency components of electromyographic signals at different times can be extracted through time-frequency analysis methods, which helps to identify muscle contraction patterns and fatigue conditions.
[0133] The above technical solution enables effective feature extraction of electromyographic signals on the lower limb surface.
[0134] Alternatively, the single-modal feature acquisition unit is configured to perform feature extraction on the lower limb inertial measurement signal when the single-modal data is a lower limb inertial measurement signal, so as to obtain single-modal features that characterize at least one of the swing amplitude, swing rhythm, stability, abnormal conditions and tilt conditions of the target object's gait.
[0135] For example, an IMU sensor can measure the acceleration, angular velocity, and attitude of a target object using an accelerometer and a gyroscope. Feature extraction schemes may include: angular velocity: measuring the rotational angular velocity of the lower limbs during movement using a gyroscope to analyze the swing amplitude and rhythm of the gait; linear acceleration: used to detect the acceleration and deceleration of the lower limbs during walking, helping to determine gait stability; attitude estimation: calculating the attitude changes of the lower limbs by fusing accelerometer and gyroscope measurements to analyze abnormalities or tilting phenomena in the gait.
[0136] The above technical solution enables effective feature extraction from lower limb IMU signals.
[0137] Alternatively, the single-modal feature acquisition unit is configured to perform feature extraction on the plantar pressure signal when the single-modal data is a plantar pressure signal, so as to obtain single-modal features that characterize the stability of the target object's gait.
[0138] For example, the plantar pressure measurement unit sensor can detect pressure changes in different areas of the foot during gait. This example can analyze the pressure distribution in the forefoot, midfoot, and hindfoot to determine if there are any abnormal force patterns in the target object, thus helping to assess walking stability; and determine the plantar contact area based on the pressure distribution to assess the stability of the gait.
[0139] The above technical solution enables effective feature extraction of plantar pressure signals.
[0140] To better understand the overall working process of the lower limb rehabilitation walking system, the following example, shown in Figure 8, will be used as an illustration. See Figure 8 for an example:
[0141] To achieve adaptive weight reduction using a multimodal lower limb rehabilitation walker, this example collects the target subject's actual weight, sEMG signals, lower limb IMU signals, and plantar pressure signals. These signals are then processed through feature extraction and fusion, and transmitted to a host computer, which sends control commands to the lower limb rehabilitation walker. The walker then uses its built-in torque motor and force feedback motor, along with PID control, to adjust the opening and closing height of the walker in real time to achieve adaptive weight reduction.
[0142] The PID control system adjusts the motor output to ensure that the support force provided by the lower limb rehabilitation walker matches the target individual's real-time movement state. For example, the support force needs to be increased when the gait is unstable, while it is gradually reduced when muscle activity is normal. The PID control calculates the error between the current motor output support force and the target weight reduction, and calculates new motor output adjustment parameters based on this error, adjusting the motor output accordingly. Furthermore, the force feedback motor can detect the target individual's real-time load; if the pressure applied by the target individual exceeds the set range, the PID control can automatically increase weight reduction support.
[0143] In addition, the system can continuously monitor the motion state and support status of the target object, and dynamically adjust the parameters of the PID control based on the real-time monitoring results, thereby optimizing the weight reduction effect.
[0144] The specific steps of adaptive weight loss are as follows:
[0145] Initialization: The system sets the initial weight-reduction support based on the target subject's actual weight and training goals. Real-time monitoring and data acquisition: Multimodal data is acquired in real time, and the target subject's movement status is analyzed through multimodal data fusion. PID control adjustment: Through PID control, the motor output is adjusted according to real-time errors to dynamically adjust the weight-reduction intensity. Status feedback and adjustment: If the target subject is detected to have gait instability, muscle fatigue, or abnormal force feedback, the PID control can quickly adjust the support intensity to ensure the target subject's safety and provide more precise support to the target subject through the force feedback motor. Training completion and summary: At the end of each training session, the system can record the target subject's movement data and weight-reduction support status, generate a report, and adjust the initial settings for the next training session, thereby ensuring personalized lower limb rehabilitation training.
[0146] The above example demonstrates how a lower limb rehabilitation walker can achieve adaptive weight reduction by combining multimodal data. Furthermore, based on the characteristics of adaptive weight reduction, the walker also features personalized training program generation, long-term data evaluation, and a multi-level safety feedback mechanism. This completely revolutionizes the way lower limb rehabilitation training is conducted, thereby greatly improving the accuracy, safety, and effectiveness of training for the target population.
Claims
1. A lower limb rehabilitation assistive system, comprising: Lower limb rehabilitation walking aids; among them, The lower limb rehabilitation walking aid includes a movable device that can move on the ground, a first adjustment device, a second adjustment device, a human body fixation device, and a main control device. One end of the movable device is V-shapedly rotatably connected to the first adjustment device and the other end is V-shapedly rotatably connected to the second adjustment device. The first adjustment device and the second adjustment device are X-shapedly rotatably connected. The first adjustment device is equipped with a first torque motor and a force feedback motor. The force feedback motor is rigidly connected to the human body fixation device. The second adjustment device is equipped with a second torque motor. The main control device includes a pressure acquisition unit, a target weight reduction acquisition unit, a motor output adjustment parameter acquisition unit, and a torque motor output adjustment unit; wherein... The pressure acquisition unit is configured to acquire the pressure applied to the human body fixation device by the target object undergoing lower limb rehabilitation training while wearing the human body fixation device via the force feedback motor. The target weight reduction obtaining unit is configured to obtain the target weight reduction of the target object; The motor output adjustment parameter obtaining unit is configured to perform proportional-integral-derivative control based on the error between the pressure and the target weight reduction to obtain the motor output adjustment parameters. The torque motor output adjustment unit is configured to adjust the output of at least one of the first torque motor and the second torque motor according to the motor output adjustment parameters, so as to adjust the opening angle between the first adjustment device and the second adjustment device, wherein the opening angle is negatively correlated with the supporting force provided to the target object by the human body fixation device, and the supporting force is used to resist the gravity of the target object.
2. The system according to claim 1, wherein, The main control device also includes a force feedback motor output adjustment unit: wherein... The force feedback motor output adjustment unit is configured to adjust the output of the force feedback motor according to the motor output adjustment parameters, so as to control the human body fixation device to rotate around the target object.
3. The system according to claim 1, further comprising: host computer; among which, The host computer includes: a control command sending unit; wherein... The control command sending unit is configured to send control commands to the lower limb rehabilitation walking aid; The target weight reduction obtaining unit is configured to analyze the received control command in response to obtain the target weight reduction of the target object.
4. The system according to claim 3, wherein, The host computer further includes: a data acquisition unit and a control command generation unit; wherein... The data acquisition unit is configured to acquire the training data of the target object; The control command generation unit is configured to determine the motion state of the target object based on the training data, and generate the control command based on the motion state.
5. The system according to claim 4, wherein, The training data is multimodal data, which includes at least two types of unimodal data; The control command generation unit includes a fusion subunit and a control command generation subunit; wherein... The fusion subunit is configured to fuse with the at least two types of single-modal data to obtain a fusion result; The control command generation subunit is configured to determine the motion state of the target object based on the fusion result, and generate the control command based on the motion state.
6. The system according to claim 5, wherein, The fusion subunit includes a single-modal feature acquisition unit and a feature fusion unit; wherein... The single-modal feature acquisition unit is configured to extract features from each of at least two types of single-modal data to obtain single-modal features; The feature fusion unit is configured to fuse all single-modal features corresponding to at least two types of single-modal data to obtain a fusion result.
7. The system according to claim 6, wherein, The single-modal feature acquisition unit is configured to perform at least one of the following operations: In the case where the single-modal data is lower limb surface electromyography (EMG) signal, feature extraction is performed on the lower limb surface EMG signal to obtain single-modal features that characterize at least one of the activation degree and fatigue status of the lower limb muscles of the target object. When the single-modal data is a lower limb inertial measurement signal, feature extraction is performed on the lower limb inertial measurement signal to obtain single-modal features that characterize at least one of the following: swing amplitude, swing rhythm, stability, abnormal conditions, and tilt conditions of the gait of the target object. as well as When the single-modal data is a plantar pressure signal, feature extraction is performed on the plantar pressure signal to obtain single-modal features characterizing the stability of the target object's gait.
8. The system according to claim 4, wherein, When the motion state indicates that the target object has a tendency to fall, the control commands include weight adjustment sub-commands and support force adjustment sub-commands, and the main control device further includes a support force adjustment unit; wherein... The target weight reduction obtaining unit is configured to analyze the received load adjustment sub-instruction in response to obtain the target weight reduction of the target object. The support force adjustment unit is configured to adjust the output of the force feedback motor in response to the received support force adjustment sub-command, so as to control the human body fixation device to rotate around the target object.
9. The system according to claim 4, wherein, The host computer also includes: a personalized training scheme acquisition unit; wherein... The personalized training scheme acquisition unit is configured to acquire the personalized training scheme of the target object; The control command generation unit is configured to determine the motion state of the target object based on the training data, and generate the control command based on the motion state and the personalized training scheme.
10. The system according to claim 9, wherein, The host computer also includes: a rehabilitation training record acquisition unit and a personalized training plan generation unit; The rehabilitation training record acquisition unit is configured to acquire the rehabilitation training record of the target object during the rehabilitation training process using the lower limb rehabilitation walking aid; The personalized training program generation unit is configured to generate the personalized training program based on the rehabilitation training record.