A wheeled active safety protection rehabilitation walking-assisting robot and its use method

Abstract

The present invention discloses a wheeled active safety protection rehabilitation walking aid robot and its use method. The robot comprises an omnidirectional motion device, a sensor detection system, and a safety protection response system. The data detected by the sensor detection system is transmitted to an industrial control computer for processing and judgment, and the safety protection response system responds accordingly based on the judgment result. The intelligent walking aid of the present invention can detect and judge the user's movement intentions, flexibly and omnidirectionally assist the user's movement, simultaneously identify environmental obstacle information and perform obstacle avoidance motion planning. Furthermore, the robot can determine the stability of the user's movement state, detect falling tendencies, and control the walking aid's steering and forward response, ensuring stability and safety for the user within a limited time.

Description

Technical Field

[0001] The present invention relates to the technology of rehabilitation walking-assist robots, and in particular to a wheeled active safety protection rehabilitation walking-assist robot and a method of using the same. Background Art

[0002] China has one of the fastest-growing aging populations in the world. By 2050, the number of elderly people is projected to exceed 300 million, accounting for one-fifth of the total population, marking the country's entry into a period of extreme aging. This aging population poses significant challenges to the economy, healthcare, and elderly care services. As the elderly's physical function gradually declines, the risk of falls increases, significantly increasing the demand for assisted mobility. Whether for mobility at home or outdoors, adequate medical care and rehabilitation robots are urgently needed resources and initiatives.

[0003] Existing technologies include a variety of rehabilitation robots for assisting walking, including exoskeleton-type rehabilitation robots and various mobile walkers. Exoskeleton-type rehabilitation robots are suitable for elderly people or patients with severely weakened lower limb muscles and inability to walk; walker-type mobile walkers are suitable for elderly people or patients with weak lower limbs and only partial walking ability; and cane-type intelligent rehabilitation robots are suitable for elderly people or patients with basic walking ability who still require support. However, these existing devices have limited functionality, primarily providing mobility assistance, lack a high degree of intelligence, and lack active protection features. Summary of the Invention

[0004] Purpose of the invention: The purpose of the present invention is to provide a safe, stable and intelligent wheeled active safety protection rehabilitation walking robot and its use method.

[0005] Technical solution: On the one hand, the present invention provides the wheeled active safety protection rehabilitation walking robot, which includes an omnidirectional motion device, a sensor detection system and a safety protection response system. The data detected by the sensor detection system is transmitted to an industrial computer for processing and judgment, and the corresponding safety protection response system responds according to the judgment result. The omnidirectional motion device includes a multi-layer walking frame and a motion controller. The multi-layer walking frame includes an upper walking frame and a lower walking frame, and the two are connected by a height adjustment device. The sensor detection system is installed on the multi-layer walking frame.

[0006] Furthermore, the height adjustment device connects the upper walker and the lower walker by means of a retractable hydraulic column and a slide rail, and the retractable hydraulic rod can adapt to different heights of users.

[0007] Furthermore, the motion controller includes a front-end Mecanum wheel, a rear-end power wheel, an industrial computer, and a motor, and the industrial computer is arranged at the bottom of the lower walking frame.

[0008] Furthermore, a Mecanum wheel is distributed on each side of the front bottom section of the lower walker, and a rear power wheel is distributed on each side of the rear bottom section of the upper walker. The motor is connected to the lower walker through a motor fixing frame, and the motor and the rear power wheel are connected through bearings in the motor fixing frame sleeve. The industrial control is placed at the bottom of the walker.

[0009] Furthermore, the sensing detection system includes a laser radar, a force sensor, a waist laser rangefinder, and a leg laser rangefinder. The laser radar is located at the front of the walker and is used to detect environmental obstacle information. The leg laser rangefinder is located on the lower walker frame and is used to monitor the forward and backward movement of the user's lower limbs. The walker waist laser rangefinder is placed on the upper walker frame and is used to detect the displacement and status of the user's waist. The force sensor is located on the upper walker frame and connected to the handrail. The laser radar, force sensor, waist laser rangefinder, and leg laser rangefinder are connected to the industrial computer via a high-speed USB interface. The leg monitoring laser rangefinder is used to detect the user's leg movement data. The waist laser rangefinder is used to detect the displacement and status of the user's waist. The laser radar is used to detect environmental obstacle information.

[0010] Furthermore, the force sensor is a thin film pressure sensor, which constitutes a force sensor matrix handle. Each force sensor matrix handle is composed of 8 thin film pressure sensors. The 8 thin film pressure sensors are distributed on the walker handle in the form of even spacing, and are used to detect the force data of the user's upper limbs to obtain the user's movement intention.

[0011] Furthermore, the smart walker has four handles, which can be used in two modes: prone mode and pushing mode. The prone mode means that the user lies on the body of the walker with the upper body and uses the front prone armrest; the pushing mode is similar to a cart and uses the rear pushing armrest.

[0012] The safety protection response system includes steering protection and forward protection. When it detects that the user is going uphill, it provides assistance in going uphill. When it detects that the user is going downhill, it actively controls the speed to prevent the speed from dropping too quickly. When it detects a fall, it moves in the opposite direction of the falling trend and brakes in time.

[0013] The industrial computer obtains the user's movement intention and movement status based on the detected force data and leg movement data, thereby assisting the user in walking and monitoring the user's movement status. It has a fall prediction function and controls the active safety protection response algorithm to provide anti-fall assistance to the user.

[0014] Another aspect of the present invention provides a method for using the wheeled active safety protection rehabilitation walker robot. The sensing detection system uses a laser radar, a force sensor, and four laser rangefinders to monitor the user's force data, environmental obstacle information, and leg motion information in real time. This information is transmitted to an industrial computer via a high-speed USB interface. If the user is in a normal state and there are no obstacles in the surrounding environment, the industrial computer uses the obstacle information and leg motion information to transmit this information to the industrial computer via the high-speed USB interface. The industrial computer uses the force data to determine the user's movement intention and transmits this movement intention to the omnidirectional walker, causing the motor to drive the powered wheels to move, thereby allowing the robot to adapt to the user's movement and provide assistance to the user during walking. If the user is in an abnormal state and obstacles are detected in the surrounding environment, the industrial computer uses an obstacle avoidance motion control algorithm to perform motion planning to guide the user to avoid the obstacles. If the user is detected ascending an incline, the robot provides assistance for ascending the incline. If the user is detected descending an incline, the robot actively controls the speed to prevent a rapid decrease in speed. If the user is detected to be in a state of falling, the industrial computer executes a fall prevention control algorithm to control the response movement, moving in the opposite direction of the falling trend and applying brakes in a timely manner, thereby preventing the movement trend and ensuring the user's stability and safety.

[0015] The specific implementation of the safety protection control algorithm is as follows: For steering protection, the walker's steering characteristics are classified into five categories: large left turn, small left turn, straight ahead, small right turn, and large right turn, based on obstacle distance information detected by the lidar and the force interaction between the user and the robot. When no obstacle is detected or the difference in the steering force between the user's two handlebars approaches zero, the desired steering behavior tends to approach zero. As the obstacle approaches or the difference in the user's steering force gradually increases, the desired steering behavior accelerates. When the obstacle distance or the steering force difference reaches a certain range, the desired steering behavior reaches a critical state and stops increasing, and the walker will turn at its built-in maximum steering angular velocity. For fall protection, the walker's upper limb force signals are detected using a force sensor matrix on the handlebars, and the lower limb motion signals are detected using a laser ranging sensor. Multi-sensor data fusion is performed using a Kalman filter algorithm to analyze the user's intended velocity. A modified sequential probability ratio test is used to detect falls using the fused data. If it is detected that the user has a tendency to fall, the user and the robot will move in the opposite direction according to the desired position of the stable state to prevent the falling tendency until the stable state is reached and brake, thereby forming a buffering movement until the user and the robot return to the stable state.

[0016] Furthermore, the fall detection method determines whether a fall has occurred based on an improved sequential probability ratio test algorithm, by adding preconditions and weights to the sequential probability ratio test, wherein the handle is completely released or the instantaneous force increases as a precondition, and the precondition is the sum of the force values ​​of the left and right handles, and a higher weight is given to the upper limbs to optimize the decision function. When the decision function Δ(k) is greater than or equal to the lower limit threshold lnB, no correction is made. When it is less than lnB, the decision function Δ(k) is corrected to Δ′(k), and negative values ​​are not accumulated, further improving the detection delay and reducing the false positive rate.

[0017] Compared with the prior art, the present invention has the following beneficial effects:

[0018] The thin-film pressure sensor matrix handle, designed specifically for the intelligent walker, enables multi-directional force signal sensing of the walker by the user, allowing the user to control the walker's forward movement, steering, and other movements through force sensors, significantly reducing the user's learning curve. The specially designed force sensor matrix handle can continuously adapt to the user's usage habits over the course of extended use, making it "more and more comfortable with use." It implements real-time monitoring and proactive protection against user falls, ensuring the safety of the intelligent walker. It can assist and protect users with gait instability, while also providing torso fixation and protection, ensuring both safety and mobility assistance. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a front view of a wheeled active safety protection rehabilitation walking-assistance robot according to the present invention;

[0020] Figure 2 This is a rear view of a wheeled active safety protection rehabilitation walking-assistance robot according to the present invention;

[0021] Figure 3 A top view of a wheeled active safety protection rehabilitation walking-assistance robot according to the present invention;

[0022] Figure 4 This is a roadmap for the overall implementation of sensing detection and safety protection for a wheeled active safety protection rehabilitation walking robot of the present invention;

[0023] Figure 5 This is a flowchart of the obstacle avoidance and protection of a wheeled active safety protection rehabilitation walking-assist robot of the present invention;

[0024] Figure 6 This is a flow chart of active steering protection for a wheeled active safety protection rehabilitation walking-assist robot according to the present invention;

[0025] Figure 7 This is a fall protection flow chart of a wheeled active safety protection rehabilitation walking assistance robot according to the present invention. DETAILED DESCRIPTION

[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0027] The wheeled active safety protection rehabilitation walking robot of this embodiment includes an omnidirectional motion device, a sensor detection system, and a safety protection response system. The omnidirectional motion device includes a multi-layered walking frame and a motion controller. Data detected by the sensor detection system is transmitted to the industrial control computer 23 for processing and evaluation, and the safety protection response system responds accordingly based on the evaluation results.

[0028] The multi-layer walker frame includes an upper walker frame 11, a bottom walker frame 12, two prone armrests 13, two pushing armrests 14, and a height adjustment device 15; the height adjustment device 15 connects the upper walker frame 11 and the bottom walker frame 12 by a retractable hydraulic column and a slide rail, and the retractable hydraulic rod can adapt to users of different heights; the intelligent walker has two usage modes: prone mode and pushing mode. The prone mode means that the user's upper body lies on the body of the walker, and the user uses the prone armrest 13 at the front; the pushing mode is similar to a cart, and the user uses the two pushing armrests 14 at the back.

[0029] The motion controller includes a front-end Mecanum wheel 21, a rear-end power wheel 22, an industrial computer 23, and a motor 24. The Mecanum wheel 21 and the rear-end power wheel 22 are evenly distributed at the bottom of the walker. The power wheel motor 24 is connected to the walker through a motor mounting bracket. The motor 24 and the power wheel are connected through bearings in the motor mounting bracket sleeve. The industrial computer 23 is placed at the front end of the bottom of the walker. Figure 1 、 2 , as shown in 3.

[0030] The sensing detection system includes a laser radar 31, a force sensor 32, a waist laser rangefinder 33 and a leg laser rangefinder 34. The safety protection response system includes steering protection and forward protection. The laser radar 31 is placed in front of the walker to detect environmental obstacle information; the leg laser rangefinder 34 is set above the lower walker 12 to detect the user's leg movement data; the waist laser rangefinder 33 is placed below the upper walker 11 to detect the displacement and state of the user's waist; the force sensor 32 is connected to the armrests 13 and 14 to detect the force data of the user's upper limbs to obtain the user's movement intention; the laser radar 31, force sensor 32 and laser rangefinders 33 and 34 are connected to the industrial computer 23 through a high-speed USB interface. The industrial computer 23 calculates the user's movement intention and movement state based on the detected force information and leg movement data, and has a fall prediction function, thereby assisting the user in walking and monitoring the user's movement state. Figure 1 、 2 , as shown in 3.

[0031] The safety protection response system includes steering protection and forward protection to provide anti-fall assistance to the user. The sensing detection system monitors the user's force data, environmental obstacle information and leg movement information in real time through a laser radar 31, a force sensor 32 and four laser rangefinders 33, 34, and transmits this information to the industrial computer 23 through a high-speed USB interface. If the user is in a normal state and there are no obstacles around the environment, the industrial computer 23 obtains the user's movement intention through the force data and transmits the movement intention to the omnidirectional motion device. The motor drives the power wheel to move, so that the robot follows the user's movement and provides assistance to the user when walking; if the user is in an abnormal state and obstacles are detected around the environment, the industrial computer 23 performs movement steering through the obstacle avoidance and steering motion control algorithm to guide the user to avoid obstacles; if it is detected that the user is going uphill, it implements assistance to go uphill, and when it is detected that the user is going downhill, it actively controls the speed to prevent the speed from dropping too quickly; if it is detected that the user has a tendency to fall, the industrial computer 23 will execute the fall protection control algorithm to control the response movement, move in the opposite direction of the falling trend and brake in time, thereby stopping the movement trend and ensuring the stability and safety of the user. Figure 4 shown.

[0032] The specific implementation of the obstacle avoidance steering motion control algorithm is as follows: according to the environmental information of the laser radar 31, a detection area is set, and the rectangular condition is used to determine whether there is an obstacle. When any scanning point in the scanning points of the forward laser radar meets the rectangular condition, it is determined that there is an obstacle in the rectangular area. Five position area judgment conditions are set and used as the signal input for the displacement control motor's direction of travel; the obstacle avoidance steering characteristics of the walker are divided into five labels, namely, large left turn, small left turn, straight, small right turn and large right turn. When no obstacle information is detected, its steering expectation is more likely to approach zero; and as the obstacle distance slowly decreases, the steering expectation begins to accelerate; when the target distance reaches a certain range, the steering expectation has entered a critical state and no longer increases, and the walker will turn at the built-in maximum steering angular velocity; such as Figure 5 shown.

[0033] The specific implementation of the active steering motion control algorithm is as follows: based on the force interaction information between the user and the robot, the force sensor matrix handle is used to use the vector superposition method to decompose the force applied to the walker handle into two directions, horizontal and vertical, and use it as the signal input for the force control motor's direction of travel; the walker's steering characteristics are divided into 5 labels, namely, large left turn, small left turn, straight, small right turn and large right turn. When the steering force difference between the two handles of the user is close to zero, the steering expectation is more likely to approach zero; and as the steering force difference slowly increases, the steering expectation begins to accelerate; when the steering force difference reaches a certain range, the steering expectation has entered a critical state and no longer increases. The walker will turn at the built-in maximum steering angular velocity, such as Figure 6 shown.

[0034] The specific implementation of the fall prevention control algorithm is as follows: the force sensor matrix on the walker handle (8 on each handle, 16 in total) is used to detect the user's upper limb force signal, and four laser rangefinders (including two waist laser rangefinders and two leg laser rangefinders) are used to detect the user's lower limb motion signal. The multi-sensor data is fused through the Kalman filter algorithm, and the improved sequential probability ratio test method (PLT-SPRT) is used to determine whether the user has fallen. If the user is detected to have a tendency to fall, the user and the robot are prevented from falling by moving in the opposite direction according to the expected position of the stable state until a stable state is reached and braking is performed, thereby forming a buffering movement until the user and the robot return to a stable state, such as Figure 5 shown.

[0035] The improved sequential probability ratio test algorithm (PLT-SPRT) is specifically implemented as follows: the sequential probability ratio test (SPRT) approximately satisfies: After taking the logarithm, the threshold becomes ln A , ln B, the decision relationship of the SPRT method is: In fall detection, the original hypothesis is normal walking, the alternative hypothesis is falling, and when the value of the judgment function is greater than the threshold ln A When the value of the judgment function is less than the threshold ln B When , the test is stopped and the normal state judgment is given. For the fall detection of the walker, the optimization is less than the threshold ln B The decision function is redefined as follows:

[0036] Where k is the sampling sequence number, when the decision function Δ(k) is greater than or equal to the lower limit threshold ln B No correction is made when it is less than ln B When Δ(k) is modified to Δ'(k), negative values ​​will not be accumulated.

[0037] In addition, preconditions and weights are added to the SPRT test, giving a higher weight to the upper limbs. The precondition here is the sum of the left and right handle force values. The optimization decision function can be further expressed as:

[0038] Δ″(k)=Δ′(k)C+ln A (1-C)

[0039]

[0040] Among them, C is the state of judging the occurrence of the precondition, r i To determine the state of the i-th precondition, that is, the condition that the handle is completely released or the instantaneous force increases significantly, there is When the decision function Δ(k) is greater than or equal to the lower threshold ln B When it is less than ln B When , the decision function Δ(k) is modified to Δ′(k), and negative values ​​will not be accumulated. Figure 7 shown.

[0041] The above description is merely a preferred embodiment of the present invention and does not limit the present invention in any way. Any person skilled in the art who, without departing from the scope of the present invention, makes any equivalent substitution, modification, or other changes to the technical solution and technical content disclosed in the present invention shall be deemed to be within the scope of the present invention and still fall within the scope of protection of the present invention.

Claims

1. A wheeled active safety protection rehabilitation walking robot, characterized in that: The invention comprises an omnidirectional motion device, a sensor detection system and a safety protection response system. The data detected by the sensor detection system is transmitted to an industrial control computer (23) for processing and judgment, and the safety protection response system responds to the corresponding safety protection response system according to the judgment result. The omnidirectional motion device comprises a multi-layer walking frame and a motion controller. The multi-layer walking frame comprises an upper walking frame (11) and a lower walking frame (12), which are connected by a height adjustment device (15). The sensor detection system is installed on the multi-layer walking frame. The fall prevention control algorithm is implemented as follows: the force sensor matrix on the walker handle detects the user's upper limb force signal, the laser ranging sensor detects the user's lower limb motion signal, and the multi-sensor data is fused through the Kalman filter algorithm. The improved sequential probability ratio test method is used on the fused data to detect whether the user has fallen. If the user is detected to have a falling trend, the reverse motion is calculated based on the expected position of the user and the robot in the stable state based on the finite time control algorithm to prevent the falling trend until the expected speed of the stable state is reached, thereby forming a buffering motion until the user and the robot return to a stable state. The improved sequential probability ratio test algorithm PLT-SPRT is specifically implemented as follows: the sequential probability ratio test (SPRT) approximately satisfies: , after taking the logarithm, the threshold becomes , , the judgment relationship of the SPRT method is: In fall detection, the original hypothesis is normal walking, and the alternative hypothesis is fall. When the value of the judgment function is greater than the threshold When the value of the judgment function is less than the threshold, the test is stopped and a fall judgment is given; when the value of the judgment function is less than the threshold When , the test is stopped and a judgment of normal state is given. For the fall detection of the walker, the optimization is less than the threshold To reduce the judgment of the normal state and further eliminate the detection delay, the optimization decision function is redefined as: , where 𝑘 is the sampling sequence number, when the decision function Greater than or equal to the lower threshold No correction is made when it is less than The decision function Corrected to , negative values ​​will not be accumulated, In addition, preconditions and weights are added to the SPRT test, giving a higher weight to the upper limbs. The precondition here is the sum of the left and right handle force values. The optimization decision function can be further expressed as: ； in, To determine the state of the precondition, To judge the The state in which the precondition occurs, that is, the handle is completely released or the force increases sharply instantly, , when the decision function 𝛥(𝑘) is greater than or equal to the lower threshold When it is less than When the decision function Corrected to , negative values ​​will not be accumulated.

2. The wheeled active safety protection rehabilitation walking-assist robot according to claim 1, characterized in that: The motion controller comprises a front-end Mecanum wheel (21), a rear-end power wheel (22), an industrial control computer (23), and a motor (24). The industrial control computer (23) is arranged at the bottom of the lower walking frame (12).

3. The wheeled active safety protection rehabilitation walking-assist robot according to claim 2, characterized in that: The sensing detection system includes a laser radar (31), a force sensor (32), a waist laser rangefinder (33) and a leg laser rangefinder (34), wherein the laser radar (31) is arranged at the front of the walker for detecting environmental obstacle information, the leg laser rangefinder (34) is arranged above the lower walker frame (12) for monitoring the forward and backward movement of the user's lower limbs, the walker waist laser rangefinder (33) is placed below the upper walker frame (11) for detecting the displacement and state of the user's waist, the force sensor (32) is arranged on the upper walker frame (11) and connected to the handrail, and the laser radar (31), the force sensor (32), the waist laser rangefinder (33) and the leg laser rangefinder (34) are connected to the industrial control computer (23).

4. The wheeled active safety protection rehabilitation walking-assist robot according to claim 3, characterized in that: Two Mecanum wheels (21) are evenly distributed at the front end of the bottom of the lower walking frame (12), and two power wheels (22) are evenly distributed at the rear end. The motor (24) is connected to the lower walking frame (12) and the Mecanum wheels (21), respectively.

5. The wheeled active safety protection rehabilitation walking-assist robot according to claim 1, characterized in that: Four handles are provided above the upper walking frame (11), including two prone armrests (13) and two pushing armrests (14), and the four handles are integrally formed.

6. The wheeled active safety protection rehabilitation walking-assist robot according to claim 5, characterized in that: The four handles detect force data through eight force sensors (32) distributed thereon, and the eight force sensors (32) form a sensor matrix.

7. The method for using the wheeled active safety protection rehabilitation walking-assist robot according to any one of claims 1 to 6, characterized in that: The sensing detection system monitors the user's force data, environmental obstacle information, and leg movement information in real time through a force sensor (32), a laser radar (31), a waist laser rangefinder (33), and a leg laser rangefinder (34), and transmits the information to an industrial computer (23) through a high-speed USB interface. If the user is in a normal state and there are no obstacles around the environment, the industrial computer (23) obtains the user's movement intention through the force data and transmits the movement intention to the omnidirectional motion device. Then, the motor (24) drives the Mecanum wheel (21) to move and perform active steering movement. If the user is in an abnormal state and obstacles are detected in the surrounding environment, the industrial computer (23) performs motion planning through an obstacle avoidance motion control algorithm to guide the user to avoid the obstacles. If the user is detected to have a tendency to fall, the industrial computer (23) will execute a fall protection control algorithm to control the response motion, move in the opposite direction of the falling tendency and brake in time, thereby preventing the motion tendency and ensuring the stability and safety of the user.

8. The method for using the wheeled active safety protection rehabilitation walking robot according to claim 7, wherein the obstacle avoidance motion control algorithm is implemented as follows: according to the environmental information of the laser radar (31), a detection area is set, and a rectangular condition is used to determine whether an obstacle exists. When any scanning point among the scanning points of the forward laser rangefinder satisfies the rectangular condition, it is determined that there is an obstacle in the rectangular area, and 5 position area determination conditions are set, and used as the signal input for the direction of travel of the displacement control motor; the obstacle avoidance steering feature of the walker is divided into 5 labels, namely, large left turn, small left turn, straight, small right turn and large right turn. When no obstacle information is detected, its steering expectation is more likely to approach zero; and as the obstacle distance gradually decreases, the steering expectation begins to accelerate; when the target distance reaches a certain range, the steering expectation has entered a critical state and no longer increases, and the walker will turn at the built-in maximum steering angular velocity.

9. The method for using the wheeled active safety protection rehabilitation walking aid robot according to claim 7, wherein the control algorithm for the active steering motion is implemented as follows: based on the force interaction information between the user and the robot, a vector superposition method is used for the force sensor matrix handle to decompose the force applied to the walker handle into horizontal and vertical directions, and use them as signal input for the force control motor's direction of travel; the steering characteristics of the walker are divided into 5 labels, namely, large left turn, small left turn, straight going, small right turn and large right turn. When the difference in steering force between the two handles of the user is close to zero, the steering expectation is more likely to approach zero; and as the difference in steering force slowly increases, the steering expectation begins to accelerate; when the difference in steering force reaches a certain range, the steering expectation has entered a critical state and no longer increases, and the walker will turn at the built-in maximum steering angular velocity.

Images