CN114770478A

CN114770478A - Remote variable-stiffness reconfigurable modular exoskeleton and control system and control method thereof

Info

Publication number: CN114770478A
Application number: CN202210538489.2A
Authority: CN
Inventors: 吴青聪; 赵子越; 朱杨辉; 陈志贤; 张烨虹; 竺文帆; 鲁嵩山
Original assignee: Nanjing University of Aeronautics and Astronautics
Current assignee: Nanjing University of Aeronautics and Astronautics
Priority date: 2022-05-18
Filing date: 2022-05-18
Publication date: 2022-07-22
Anticipated expiration: 2042-05-18
Also published as: CN114770478B

Abstract

The invention discloses a far-end variable-stiffness reconfigurable modular exoskeleton which comprises a wearing mechanism, a plurality of far-end variable-stiffness drivers and a plurality of exoskeleton modules, wherein the far-end variable-stiffness drivers are arranged on the wearing mechanism and comprise an input disc, an output disc and a driving motor for driving the input disc to rotate; the exoskeleton module comprises an upper connecting plate, a lower connecting plate and a lasso wheel fixedly connected with the lower connecting plate, wherein the lower connecting plate rotates relative to the upper connecting plate to drive the lasso to wind the lasso wheel and the output disc to form belt transmission; one distal variable stiffness driver corresponds to one exoskeleton module. The exoskeleton is modularized, a single module can be used independently, and customized multi-joint combination can be used, so that the requirements of various working scenes are met, the cost is lower, and the later maintenance is convenient; the utility model adopts the lasso transmission of distal end, reduces articular inertia, places the bionical driver that the quality is big at the back, accords with human atress custom, and the operation is convenient in narrow space.

Description

Remote variable-rigidity reconfigurable modular exoskeleton and control system and control method thereof

Technical Field

The invention relates to an auxiliary robot, in particular to a remote variable-stiffness reconfigurable modular exoskeleton and a control system and a control method thereof.

Background

Traditional industrial robot can replace the people to accomplish partial repetitive labor at to a great extent, has both practiced thrift the cost of labor and has also improved production efficiency, but its limitation that is difficult to break through: firstly, the robot cannot completely replace the human brain, cannot cope with various accidental and special scenes in actual production, and although the robot can be made to have a certain degree of intelligence by combining the technologies such as machine vision, artificial intelligence and the like which are currently in development momentum, the implementation to the actual application scene has considerable technical difficulty, and all actual situations cannot be considered. Secondly, human beings have various sensing abilities such as vision, sense of hearing, sense of touch, temperature, can make judgement and decision rapidly according to the information of perception simultaneously. Even if various sensors are equipped for the robot to enable the robot to have various sensing capabilities, the complexity degree related to information fusion processing is extremely high. Meanwhile, in some special operation scenes, an industrial robot with a complex mechanism cannot perform flexible operation like a human. Therefore, human-machine co-fusion cooperation of the human-exoskeleton system becomes a hot technology of intelligent robot research.

Traditional exoskeletons are rigidly driven, and in many practical operation scenes, an actuating mechanism is often required to contact with the environment for flexible interaction, such as: grinding, polishing, assembling and the like. Traditional rigid driving exoskeletons rely on control algorithms to achieve flexibility in operation, and complex control algorithms rely on recognition of human motion intent to perform accurate pattern recognition. However, in many industrial operation scenes, the operation environment is complex, the acquisition interference on the sensing signals is large, and the movement intention of the human body is difficult to judge. Meanwhile, the multi-sensor fusion and complex algorithm has extremely high requirements on hardware and high power consumption, and is difficult to adapt to complex industrial operation environment at present. Therefore, the exoskeleton robot with actively variable physical rigidity needs to be developed to meet the requirements of various working scenes.

Most of the existing exoskeletons for industrial operation are unpowered exoskeletons or quasi-unpowered exoskeletons, the exoskeletons can only assist workers to carry out simple operation, and although the operation burden of the workers can be reduced to a certain extent, the operation speed and quality cannot be fundamentally improved. The existing exoskeleton for mature assistance industrial operation mainly focuses on the field of carrying the exoskeleton at present, the carrying assistance exoskeleton fundamentally improves the carrying efficiency of workers and greatly reduces the burden of the workers, but the existing exoskeleton is single in operation scene and does not have good adaptability to multiple operation scenes. Therefore, there is a need to develop a modular and reconfigurable power-assisted exoskeleton coordinated multi-joint cooperative operation control algorithm to meet the needs of various industrial operation scenarios.

In chinese patent application No. CN201921702812.5, an industrial exoskeleton for upper limbs that reduces vibration impact is disclosed, which comprises a fixture for fixing a vibration tool, and a wearing piece that is worn and coupled to the upper limbs of a human body. The upper limb industrial exoskeleton is suitable for operation using a vibration tool, but the exoskeleton can only reduce vibration to a certain extent, the rigidity of the flexible mechanism cannot be adjusted, the applicable operation scenes are few, meanwhile, assistance cannot be brought to joints, and the operation efficiency cannot be fundamentally improved.

In chinese patent application No. CN202010132049.8, an adaptive compliance control method for an upper limb rehabilitation exoskeleton robot is disclosed, in which an admittance control model is used as a control outer loop, and an adaptive global fast terminal sliding mode controller is used as a control inner loop, so as to form a complete adaptive compliance control loop. The compliance control algorithm can realize the compliance during interaction through planning and controlling the motion trail, but in a complex operation environment, the compliance control method has extremely high requirements on the real-time response performance of a system and is difficult to realize.

In chinese patent application No. CN201711386675.4, a modular reconfigurable multifunctional exoskeleton robot is disclosed, which comprises a track indoor environment, a multi-degree of freedom support frame and a whole body exoskeleton device. The invention is suitable for a plurality of virtual specific application environments, and has diversified, modularized and reconfigurable functional structures. However, the driving mechanisms of the modularized reconfigurable exoskeleton are all arranged at the joints of a human body, so that the joint quality and inertia are large, and the exoskeleton does not have the variable stiffness characteristic, so that the operation comfort degree of workers can be greatly influenced in actual operation.

Disclosure of Invention

The purpose of the invention is as follows: in view of the above shortcomings, the present invention provides a remote variable stiffness reconfigurable modular exoskeleton that combines a reconfigurable exoskeleton with actively changing physical interaction stiffness.

The invention further provides a remote variable-rigidity reconfigurable modular exoskeleton control system and a control method for improving flexibility and safety of multi-operation scene operation.

The technical scheme is as follows: in order to solve the problems, the invention adopts a far-end variable-stiffness reconfigurable modular exoskeleton which comprises a wearing mechanism, a plurality of far-end variable-stiffness drivers and a plurality of exoskeleton modules, wherein the far-end variable-stiffness drivers are arranged on the wearing mechanism and comprise driving motors, speed reducers matched with the driving motors, driving rotating shafts fixedly connected with the output of the speed reducers, input disks arranged on the driving rotating shafts, output disks also arranged on the driving rotating shafts, and stiffness adjusting devices for adjusting the connection stiffness between the input disks and the output disks, and the input disks are fixed relative to the driving rotating shafts; the output disc rotates around the driving rotating shaft; a driving lasso is wound on the output disc;

the exoskeleton module comprises an upper connecting plate, a lower connecting plate arranged at the lower end of the upper connecting plate and a noose wheel fixedly connected with the upper end of the lower connecting plate, the lower connecting plate and the noose wheel rotate relative to the upper connecting plate, the driving noose is wound on the noose wheel, and the output disc, the noose wheel and the driving noose form belt transmission; the one distal variable stiffness driver corresponds to one exoskeleton module.

Furthermore, the rigidity adjusting device comprises a plurality of input pulleys arranged on the input disc and facing one side of the output disc, a plurality of output pulleys arranged on the output disc and facing one side of the input disc, and a steel wire rope used for transmitting torque between the input pulleys and the output pulleys, wherein one end of the steel wire rope is connected with one end of a variable rigidity tension spring, and the other end of the variable rigidity tension spring is fixedly connected with the input disc; the other end of the steel wire rope is fixed on a rope disc, the rope disc is positioned on the input disc through a rigidity adjusting rotating shaft, the rope disc is fixedly connected with the rigidity adjusting rotating shaft and rotates relative to the input disc, and the rigidity adjusting rotating shaft is parallel to the driving rotating shaft and penetrates through the input disc to be fixedly connected with the driven belt wheel; the rigidity adjusting device is provided with a rigidity adjusting motor, an output shaft of the rigidity adjusting motor is connected with a worm, the worm is matched with a worm wheel, a driving belt wheel is coaxially fixed on the worm wheel, a synchronous belt is sleeved on the driving belt wheel and the driven belt wheel, and the driving belt wheel rotates to drive the driven belt wheel to rotate through the synchronous belt, so that a rope disc is driven to rotate. The far-end variable-rigidity driver is driven by a motor, and the motor improves the torque through a worm gear and a worm and drives the driving rotating shaft to rotate. The drive shaft transfers torque to the input disc. The output disc is connected with the driving rotating shaft through a bearing and is positioned through a shaft sleeve, and the input disc outputs torque to the output disc through a rigidity adjusting device. The pulley blocks are uniformly distributed on one side, facing the output disc, of the input disc, the pulley blocks arranged on one side, facing the input disc, of the output disc transmit torque through a steel wire rope winding, one end of each steel wire rope, a rigidity tension spring and a tension sensor are fixed on the input disc, and the other end of each steel wire rope is wound on a rope disc arranged on the input disc. This places the stiffness adjustment device in series between the input and output discs, providing compliance to the system. The joint angular stiffness is changed by controlling the pretightening force of the steel wire rope. The shaft of the first encoder is inserted into the shaft hole of the driving rotating shaft in an interference manner, and the first encoder shell is fixed on the output disc through the sensor support and used for measuring the elastic deflection angle theta of the input disc and the output disc.

The driven pulley drives the rope pulley through adjusting the rigid rotating shaft, so that the elongation of the variable-stiffness tension spring is adjusted to realize the change of angular stiffness. The driven belt wheel forms a planetary transmission pair through a synchronous belt and a turbine-synchronous composite wheel. The turbine-synchronous composite wheel comprises a turbine and a driving belt wheel which are coaxially and fixedly connected, and the turbine-synchronous composite wheel is arranged on the driving rotating shaft through a bearing structure and is not driven by the driving rotating shaft. The rigidity adjusting motor drives the turbine part of the turbine-synchronous composite wheel through a worm rod fixedly connected with the rigidity adjusting motor, so that the turbine-synchronous composite wheel is driven, the driven belt wheel is driven through the synchronous belt, and the driven belt wheel rotates to change the rigidity of the driver. The rope disk is positioned on the input disk, so that the rope disk can rotate around the driving rotating shaft along with the rotation of the input disk, and the driven belt wheel also revolves around the driving rotating shaft.

The exoskeleton module further comprises an outer side half shaft plate, an inner side half shaft plate, a connecting support and a second encoder, the rope sleeving wheel is connected with the outer side half shaft plate and the inner side half shaft plate through bearings, the outer side half shaft plate and the inner side half shaft plate are fixedly connected through the connecting support, the rope sleeving wheel is fixedly connected with the lower connecting plate, and the upper connecting plate is fixedly connected with the inner side half shaft plate. The second encoder shell is fixed in the rope sleeving wheel through a bolt, and the second encoder shaft is inserted into the shaft hole of the inner side half shaft plate in an interference mode and used for measuring the rotation angle of the exoskeleton. Through inboard semi-axis board and the different articulated upper junction plate of bolt fastening to through the fixed different articulated protective equipment of U type groove and screw hole of reserving on the connecting plate, the connecting plate also can be through the fixed different protective equipment of U type groove and screw hole of reserving under the same principle. Only the wearing parts and the connecting plate need to be replaced for different joints. The reconfigurable module can be used independently or in combination with a plurality of modules, and the requirements of reconfigurability and modularization are met.

The sensing system comprises an encoder of a driving motor in the far-end variable stiffness driver, an encoder of a stiffness adjusting motor, a tension sensor, a first encoder and a second encoder in the exoskeleton module.

The far-end variable-rigidity driver further comprises a driver support, the driving rotating shaft is sleeved with the driving support, the driving motor is fixed on the driving support, and a through hole through which the driving lasso penetrates is formed in the upper end of the driving support.

The invention also adopts a control system of the remote variable-stiffness reconfigurable modular exoskeleton, which comprises a data acquisition module, a model establishing module, a judgment module and a control module; the data acquisition module is used for acquiring the data of the exoskeleton in real time when the exoskeleton module acts;

the model establishing module is used for establishing a human-computer interaction operation model according to the data acquired by the data acquisition module;

the judgment module is used for judging the operation state of the exoskeleton module and judging whether to control the exoskeleton module according to the operation state of the exoskeleton module;

the control module is used for controlling the distal variable stiffness driver to control the exoskeleton module.

Furthermore, the control module controls the remote variable stiffness driver by controlling the rotating speeds of the stiffness adjusting motor and the driving motor, so that the stiffness and the torque are output to the exoskeleton module, and the control on the output torque and the output stiffness is realized.

The invention also discloses a control method of the remote variable-stiffness reconfigurable modular exoskeleton, which comprises the following steps:

step S100: carrying out simulation operation actions under different operation scenes on the exoskeleton module, and simultaneously collecting exoskeleton data when the exoskeleton module acts;

step S200: establishing a man-machine interaction operation model according to the acquired data; the method comprises the steps of including a desired torque track model and a desired rigidity track model;

step S300: carrying out actual operation tasks by using the exoskeleton module, simultaneously acquiring data of the exoskeleton in real time, and judging the operation state of the exoskeleton module according to the man-machine interaction operation model;

step S400: judging whether to control the exoskeleton module according to the operation state of the exoskeleton module; if the exoskeleton module is controlled, performing step S500;

step S500: controlling the exoskeleton module by controlling the remote variable stiffness driver according to the human-computer interaction operation model and the operation state of the exoskeleton module; according to a set expected torque track model, tracking and controlling the auxiliary torque of the exoskeleton module by adopting a torque estimation model; according to a set expected rigidity track model, closed-loop PID control is adopted to realize the tracking control of the rigidity of the exoskeleton module;

step S600: repeating the step S300 to the step S500 until the actual job task is completed; collecting the data of the outer skeleton in the process of completing the operation task in real time;

step S700: and correcting the human-computer interaction operation model by using the data acquired in real time.

Further, in the step S200, inputting the collected exoskeleton data into a teaching database, then using a TP-GMM model to learn and carry data in the teaching database and generate a probability torque trajectory, simultaneously generating a theoretical torque trajectory according to an operation scene and a human body dynamics model, and performing probability density correction on the probability torque trajectory according to the theoretical torque trajectory to obtain an expected torque trajectory model; and combining the expected torque track model with the mechanical model of admittance control to obtain an expected rigidity track model.

Further, the torque control model when the distal end variable stiffness driver controls the exoskeleton module torque is as follows:

wherein, a is the radius from the pulley center of the input pulley to the rotation center of the input disc, K is the equivalent rigidity of the steel wire rope and the variable rigidity tension spring, b is the center distance of the input pulley block, c is the radius from the pulley center of the output pulley to the rotation center of the output disc, and F₀Is the pretightening force of the variable-rigidity tension spring, theta is the elastic deflection angle, F_sIn order to drive the pretightening force of the lasso, r is the radius of a lasso wheel on the exoskeleton module, mu is the friction coefficient of the lasso and the sleeve, and psi is the curvature of the lasso sleeve.

Further, the stiffness control model when the distal end variable stiffness driver controls the stiffness of the exoskeleton module is as follows:

and theta 1 is the scale of the driving motor encoder, and theta 2 is the scale of the exoskeleton module encoder.

Further, in step S400, a threshold is set for the elastic deflection angle θ of the distal variable stiffness driver according to different work tasks, and when an actual work task is performed, if the elastic deflection angle θ acquired in real time is lower than the set threshold, the distal variable stiffness driver does not perform the assisting work.

Has the beneficial effects that: compared with the prior art, the invention has the remarkable advantages that:

(1) the exoskeleton is modularized, compared with the traditional exoskeleton, the exoskeleton is designed in a reconfigurable modularization mode, the assistance of multiple joints of a human body is realized, a single module can be used independently, the applicability is better for occasions needing only single-joint assistance, the customized multi-joint combination can be used, the requirements of various working scenes can be met, the cost is lower, and the later maintenance is convenient; and the distal lasso transmission is adopted, so that the inertia of joints is reduced, the bionic driver with large mass can be placed on the back, the stress habit of a human body is better met, and the operation in a narrow space is more convenient.

(2) Compared with the traditional variable-rigidity driver, the planetary transmission pair with unique design is adopted, the rigidity adjusting motor can be fixedly connected with the driving motor on the motor fixing frame, the defect that the traditional rigidity adjusting motor needs to swing along with an input or output mechanism is overcome, and the influence of nonlinear factors such as rotational inertia, friction and the like which are not beneficial to subsequent control is greatly reduced.

(3) The theoretical torque track in operation is obtained through calculation according to an operation scene and a human body dynamics model, then a TP-GMM (task parameterized Gaussian mixture model) machine learning model and an admittance control model are combined to obtain a rigidity track through calculation, the rigidity track is continuously adjusted through an offline learning mode, and simple and efficient online control is realized through complex offline learning.

(4) The physical rigidity of the exoskeleton module is directly controlled to realize flexible control, and meanwhile, the quick adjustment of multi-joint cooperative control can be realized for a plurality of operation scenes through a customized modular control method, so that the requirement on system hardware is lower, and the development is more convenient and quicker.

Drawings

Figure 1 is a schematic diagram of a configuration of an exoskeleton module worn by a wrist joint in accordance with the present invention;

fig. 2 is a schematic diagram of a knee worn exoskeleton module according to the present invention;

FIG. 3 is a schematic diagram of the construction of the hip-worn exoskeleton module of the present invention;

FIG. 4 is a schematic diagram of the configuration of an ankle wearable exoskeleton module of the present invention;

FIG. 5 is a schematic diagram of the construction of the exoskeleton module worn by the lower extremities (knee joint and ankle joint) of the present invention;

FIG. 6 is an exploded view of the distal variable stiffness driver of the present invention;

FIG. 7 is a schematic view of the overall structure of the distal variable stiffness driver of the present invention;

FIG. 8 is a schematic diagram of the exoskeleton module of the present invention;

FIG. 9 is an exploded view of the exoskeleton module of the present invention;

FIG. 10 is a flow chart of a control method of the present invention;

FIG. 11 is a block diagram of a real-time online control method according to the present invention;

FIG. 12 is a flowchart illustrating off-line learning in the control method of the present invention.

Detailed Description

Example 1

As shown in fig. 1 to 4, in the remote variable stiffness reconfigurable modular exoskeleton of the present embodiment, a remote variable stiffness driver ii is disposed on a wearing mechanism on the back of a user i, and assists the user i to perform work by combining a multi-joint variable stiffness work control method through an exoskeleton module iii (elbow joint, knee joint, ankle joint). As shown in fig. 5, the distal variable stiffness drivers ii and iii are placed on a wearing mechanism on the back of the user i, and the knee joint configuration iv and the ankle joint configuration v form a lower limb configuration, which can be used for rehabilitation and power assistance.

As shown in FIG. 6, the distal variable stiffness driver comprises a driver bracket 19, the driver bracket 19 is connected with a motor bracket 18 through a bolt, a driving motor 10 and a stiffness adjusting motor 28 are fixed on the motor bracket 18 through bolts, two openings are formed in two sides of the lower end of the driver bracket 19 and used for driving the rotating shaft 14 to pass through, and an opening is formed in the upper end of the driver bracket 19 and used for driving the lasso 32 to pass through. An output shaft of the driving motor 10 is fixedly connected with a worm 9 of the speed reducer, the extending direction of the driving motor 10 is parallel to the extending direction of the worm 9, the worm wheel 6 is matched with the worm 9, and the worm wheel 6 is fixedly sleeved on the driving rotating shaft 14 through a set screw to drive the driving rotating shaft 14 to rotate. The drive shaft 14 is positioned on the drive support 19 by a deep groove ball bearing 13 and is limited by the support front end cover 8 and the support rear end cover 11. The input disc 1 is fixedly sleeved on the driving rotating shaft 14, the driving rotating shaft 14 and the input disc 1 adopt a spline connection mode to transmit large torque to the input disc 1, and the input disc 1 is positioned on the driving rotating shaft 14 through a stepped shaft sleeve 34. The output disc 12 is sleeved on the driving rotating shaft 14 through a deep groove ball bearing and is positioned through shaft sleeves at two ends.

The input pulley block is fixed on the input disc 1 through a roller screw 4, and the output pulley block is fixed on the output disc 12 through a roller screw. In this embodiment, the input pulley block includes 6 input pulleys 2, the input pulleys 2 are uniformly distributed on one side of the input disc 1, and transmit torque with 6 groups of pulleys (12 output pulleys 3) on one side of the output disc 12 through wire winding of a steel wire rope 20, the arrangement relationship between the steel wire rope 20 and the input pulleys 2 and the output pulleys 3 refers to the chinese patent application with the prior art application number of 2019107682555, one end of the steel wire rope 20 is connected with one end of a variable stiffness tension spring 24, and a tension sensor 25 is respectively connected with the variable stiffness tension spring 24 and a spring support 26 through two hanging rings 27 and is fixed on the input disc 1 through the spring support 26; the other end of the steel wire rope 20 is wound on the rope reel 5, the rope reel 5 is positioned on the input disc through the rigidity adjusting rotating shaft 15, the rigidity adjusting rotating shaft 15 penetrates through the input disc 1, the other end of the rigidity adjusting rotating shaft 15 is fixedly connected with the driven belt wheel, and the driven belt wheel rotates to drive the rigidity adjusting rotating shaft 15 to rotate relative to the input disc, so that the rope reel is driven to rotate. This spring mechanism is connected in series between the input disc 1 and the output disc 12 to provide compliance to the system. The first encoder 7 is bolted to the drive carrier 19 and is connected to the output disc for measuring the angle of deflection of the input disc 1 and the output disc 12 due to the elasticity.

The driven pulley and the turbine-synchronous composite wheel 23 form a planetary transmission pair through a synchronous belt 22, the turbine-synchronous composite wheel 23 comprises a turbine and a driving pulley which are coaxially and fixedly connected, and the turbine-synchronous composite wheel 23 is sleeved on the driving rotating shaft 14 through a bearing structure and is not driven by the driving rotating shaft 14. The stiffness adjusting motor 28 drives the turbine part of the turbine-synchronous composite wheel 23 through a worm rod fixedly connected with the turbine part, so that the turbine-synchronous composite wheel 23 is driven, then the driven pulley is driven through the synchronous belt 22, so that the driven pulley rotates, the rope disc 5 is positioned on the input disc 1, and therefore the rope disc rotates around the driving rotating shaft 14 as a central shaft along with the rotation of the input disc 1, and the driven pulley also revolves around the axis of the driving rotating shaft 14 through the synchronous belt 22 so as to change the rigidity of the driver.

The drive noose 32 is sleeved over the output disc 12 through an opening in the upper end of the drive bracket 19, the drive noose 32 is also sleeved over the exoskeleton module, and the output disc 12 transmits torque to the exoskeleton module through the drive noose 32.

As shown in fig. 7 and 8, the exoskeleton module includes a lateral axle plate 35, a bearing 36, a connecting bracket 37, a fixing bracket 38, an upper connecting plate 39, a limiting block 40, an end cap 41, a second encoder 42, a lasso wheel 43, a lower connecting plate 44, a bearing 45, and a medial axle plate 46. One side of the rope sleeving wheel 43 is connected with the end cover 41 through a bolt, the other side of the rope sleeving wheel 43 is connected with the lower connecting plate through a bolt, the bearing 36 and the bearing 45 are respectively arranged on two sides of the fixing structure, the bearing 36 is arranged on a shaft shoulder of the outer side half shaft plate 35, the bearing 45 is arranged on a shaft shoulder of the inner side half shaft plate 46, and the outer side half shaft plate 35 is fixedly connected with the inner side half shaft plate 46 through the connecting support 37. The upper end of the inner side semi-axis plate 46 is fixedly connected with the fixing frame 38, and the upper part of the fixing frame 38 is fixedly connected with the upper connecting plate. The second encoder 42 housing is bolted into the noose wheel 43 and the second encoder 42 shaft is inserted into the shaft hole of the inboard axle plate for measuring the rotational angle of the exoskeleton module. Modular reconfigurable exoskeletons of different configurations are constructed by replacing different upper and

lower connector plates

39, 44 in conjunction with the above-described modular base members. The fixed frame 38 is provided with a limit block 40, and the motion range of the modular exoskeleton is limited by the limit block 40 so as to ensure safety.

Example 2

The control system of the remote variable-stiffness reconfigurable modular exoskeleton in the embodiment comprises a data acquisition module, a model establishment module, a judgment module and a control module; the data acquisition module acquires data of the exoskeleton during the action of the exoskeleton module in real time, wherein the data comprises off-line learning teaching data for performing an action test in the early stage and real-time data for performing an actual operation task; the model establishing module is used for establishing a human-computer interaction operation model according to the data acquired by the data acquisition module and continuously adjusting the interaction operation model through real-time data; the judging module is used for judging the operation state of the exoskeleton module and judging whether to control the exoskeleton module according to the operation state of the exoskeleton module; the control module controls the far-end variable stiffness driver by controlling the rotating speeds of the stiffness adjusting motor and the driving motor, outputs stiffness and torque to the exoskeleton module, and performs closed-loop PID control on the exoskeleton module to realize control on output torque and output stiffness.

Example 3

As shown in fig. 9, the control method for the remote variable stiffness reconfigurable modular exoskeleton in the present embodiment includes the following steps:

step S100: carrying out simulation operation actions under different operation scenes on the exoskeleton module, and simultaneously acquiring data of the exoskeleton when the exoskeleton module acts; as shown in fig. 10, a user first wears the exoskeleton module to perform various actions, collects teaching data according to multiple sensors, and records the teaching data into a teaching database.

Step S200: establishing a man-machine interaction operation model according to the acquired data; and (3) learning and carrying data in a teaching database by using a TP-GMM (Gaussian mixture model with parameterized tasks) to generate a probability torque track, generating a theoretical torque track according to a human body dynamics model under a specific task, and performing probability density correction on the generated probability torque track through the theoretical torque track to obtain a theoretical probability track model. And calculating to obtain a rigidity coefficient in admittance control by taking the track as a reference track and combining with a mechanical model of admittance control, wherein the rigidity coefficient track is taken as an initial expected rigidity track. The established human-computer interaction model obtains expected torque and rigidity tracks and can be called according to requirements.

Step S300: the method comprises the steps that an exoskeleton module is used for carrying out actual operation tasks, data of an exoskeleton are collected in real time, and the operation state of a user is judged in real time according to an elastic deflection angle theta in a far-end variable stiffness driver generated when a worker moves;

step S400: judging whether the torque and the rigidity of the exoskeleton module need to be adjusted to be matched with the current operation state or not by combining the torque and the rigidity reference track of the exoskeleton module; setting a threshold value for the elastic deflection angle theta of the driver according to different operation tasks, controlling the operation when the elastic deflection angle theta is lower than the threshold value, and reducing the rigidity of the driver when the elastic deflection angle theta is higher than the threshold value, so that a user can freely operate, and the use safety of the exoskeleton is ensured; if the exoskeleton module is controlled, performing step S500;

step S500: controlling the exoskeleton module by controlling a remote variable stiffness driver according to the human-computer interaction operation model and the operation state of the exoskeleton module; tracking and controlling the exoskeleton auxiliary torque by adopting a torque estimation model according to the set exoskeleton expected torque track; according to the set reference rigidity of the exoskeleton, closed-loop PID control is adopted to realize the tracking control of the exoskeleton rigidity; as shown in fig. 10, according to the torque and stiffness trajectory model of the driver, the tracking of the interactive torque and the interactive stiffness is converted into the position tracking of the elastic deflection angle, and further converted into the trajectory tracking of the driving motor and the stiffness adjusting motor in the speed mode, so that the complexity of control is greatly reduced. Meanwhile, although the accuracy of tracking the tail end track is sacrificed to a certain extent by the mode of changing the physical interaction rigidity, the comfort and the safety of the man-machine interaction are greatly improved, and meanwhile, the influence of the accuracy error on the operation can be ignored due to the existence of workers.

When the far-end variable stiffness driver controls the torque of the exoskeleton module, the torque estimation model is as follows:

wherein, a is the radius from the pulley center of the input pulley to the rotation center of the input disc, K is the equivalent rigidity of the steel wire rope and the variable rigidity tension spring, b is the center distance of the input pulley block, c is the radius from the pulley center of the output pulley to the rotation center of the output disc, and F₀The pre-tightening force of the tension spring with variable rigidity is theta, the elastic deflection angle is F_sIn order to drive the pretightening force of the lasso, r is the radius of a lasso wheel on the exoskeleton module, mu is the friction coefficient of the lasso and the sleeve, and psi is the curvature of the lasso sleeve.

When the remote rigidity-variable driver controls the rigidity of the exoskeleton module, the rigidity control model is as follows:

Step S600: repeating the steps S300 to S500 until the actual operation task is completed, thereby realizing the assistance to the operation of workers; collecting the data of the outer skeleton in the process of completing the operation task in real time; different joints can be controlled separately, and cooperative operation can be carried out by carrying out free combination aiming at different operation tasks, and the difficulty that the cooperative control of each joint needs decoupling is avoided by a modularized control mode, so that the difficulty of the integral cooperative control is greatly reduced.

Step S700: under the initial human-computer interaction operation model, data after actual operation tasks of a user are recorded into a teaching database, an expected torque track is corrected, an expected rigidity track is corrected according to comfort feedback in an operation process fed back by the user, the human-computer interaction model obtained through off-line learning is continuously corrected, the precision error problem is reduced, and meanwhile, the comfort of the user during operation is continuously improved.

Claims

1. A far-end variable-stiffness reconfigurable modular exoskeleton is characterized by comprising a wearing mechanism, a plurality of far-end variable-stiffness drivers and a plurality of exoskeleton modules, wherein the far-end variable-stiffness drivers are arranged on the wearing mechanism and comprise driving motors (10), speed reducers matched with the driving motors, driving rotating shafts (14) fixedly connected with the output of the speed reducers, input disks (1) installed on the driving rotating shafts, output disks (12) also installed on the driving rotating shafts, and stiffness adjusting devices for adjusting the connection stiffness between the input disks (1) and the output disks (2), and the input disks (1) are fixed relative to the driving rotating shafts; the output disc (12) rotates around the driving rotating shaft; a driving lasso is wound on the output disc (12);

the exoskeleton module comprises an upper connecting plate (39), a lower connecting plate (44) arranged at the lower end of the upper connecting plate (39), and a noose wheel (43) fixedly connected with the upper end of the lower connecting plate, wherein the lower connecting plate and the noose wheel rotate relative to the upper connecting plate, the driving noose is wound on the noose wheel (43), and the output disc, the noose wheel and the driving noose form belt transmission;

the one distal variable stiffness driver corresponds to one exoskeleton module.

2. The remote variable stiffness reconfigurable modular exoskeleton of claim 1, wherein the stiffness adjusting device comprises a plurality of input pulleys (2) arranged on one side of the input disc (1) facing the output disc, a plurality of output pulleys (3) arranged on one side of the output disc (12) facing the input disc, and a steel wire rope (20) used for transmitting torque between the input pulleys and the output pulleys, one end of the steel wire rope is connected with one end of a variable stiffness tension spring (24), and the other end of the variable stiffness tension spring (24) is fixedly connected with the input disc; the other end of the steel wire rope is fixed on a rope disc (5), the rope disc (5) is positioned on an input disc through a rigidity adjusting rotating shaft (15), the rope disc (5) is fixedly connected with the rigidity adjusting rotating shaft and rotates relative to the input disc, and the rigidity adjusting rotating shaft is parallel to a driving rotating shaft (14) and penetrates through the input disc to be fixedly connected with a driven belt wheel; rigidity adjusting device sets up transfers just motor (28), transfers just motor output shaft to connect the worm, and the worm cooperation has the turbine, and the turbine is coaxial to be fixed with driving pulley, and the cover is equipped with hold-in range (22) on driving pulley and the driven pulley, driving pulley rotates and drives driven pulley through the hold-in range and rotates to drive the rope dish and rotate.

3. The distal variable stiffness reconfigurable modular exoskeleton of claim 1, wherein the distal variable stiffness driver further comprises a driver bracket (19), the driving shaft (14) is sleeved with the driving bracket (19), the driving motor (10) is fixed on the driving bracket (19), and a through hole through which the driving lasso passes is arranged at the upper end of the driving bracket (19).

4. A control system for a remote variable stiffness reconfigurable modular exoskeleton as claimed in any one of claims 1 to 3 comprising a data acquisition module, a model building module, a decision module and a control module; the data acquisition module is used for acquiring the data of the exoskeleton during the action of the exoskeleton module in real time;

the model establishing module is used for establishing a man-machine interaction operation model according to the data acquired by the data acquisition module;

5. The control system of claim 4, wherein the control module controls the distal variable stiffness driver to adjust the stiffness and torque of the exoskeleton module by controlling the rotational speeds of the stiffness adjusting motor and the drive motor.

6. A control method of the control system according to claim 4 or 5, characterized by comprising the steps of:

step S100: carrying out simulation operation actions under different operation scenes on the exoskeleton module, and simultaneously acquiring data of the exoskeleton when the exoskeleton module acts;

step S200: establishing a human-computer interaction operation model according to the acquired data;

step S500: controlling the exoskeleton module by controlling a remote variable stiffness driver according to the human-computer interaction operation model and the operation state of the exoskeleton module;

step S600: repeating the step S300 to the step S500 until the actual operation task is completed; collecting the data of exoskeletons in real time in the process of completing the operation task;

step S700: and correcting the man-machine interaction operation model by using the data acquired in real time.

7. The control method according to claim 6, wherein in step S200, the collected exoskeleton data is recorded into a teaching database, and then a TP-GMM model is used to learn and carry data in the teaching database and generate a probability torque trajectory, and meanwhile, a theoretical torque trajectory is generated according to a working scene and a human body dynamics model, and a probability density correction is performed on the probability torque trajectory according to the theoretical torque trajectory to obtain an expected torque trajectory model; and combining the expected torque track model with an admittance controlled mechanical model to obtain an expected rigidity track model.

8. The control method of claim 6, wherein the torque control equation for the distal variable stiffness driver controlling exoskeleton module torque is:

9. The control method of claim 8, wherein the stiffness control equation for the distal variable stiffness driver in controlling the stiffness of the exoskeleton module is:

and theta 1 is the scale of the encoder of the driving motor, and theta 2 is the scale of the encoder of the exoskeleton module.

10. The control method according to claim 8 or 9, wherein in step S400, a threshold is set for the elastic deflection angle θ of the distal variable stiffness driver according to different job tasks, and when an actual job task is performed, if the elastic deflection angle θ collected in real time is lower than the set threshold, the distal variable stiffness driver is controlled to perform job control.