CN104698848A - Control method for rehabilitation training of lower extremity exoskeleton rehabilitation robot - Google Patents

Abstract

The invention discloses a control method for rehabilitation training of a lower extremity exoskeleton rehabilitation robot. The control method comprises the following steps: modeling the lower extremity exoskeleton rehabilitation robot to obtain a model equation, re-constructing a self-adaptive controller and a control rate calculating formula, wherein undetermined control gain parameters are obtained by pre-training; then, practically measuring rotation angles and rotation speed of knee joints and a hip joint of the lower extremity exoskeleton rehabilitation robot; finally, substituting the practical parameters into the self-adaptive control rate calculating formula to obtain a self-adaptive controller a moment t, wherein a drive device applies acting force to the lower extremity exoskeleton rehabilitation robot according to a self-adaptive control rate.

Description

A kind of control method of lower limb exoskeleton rehabilitation robot rehabilitation training

Technical field

The invention belongs to technique of medical rehabilitation field, more specifically say, relate to a kind of control method of lower limb exoskeleton rehabilitation robot rehabilitation training.

Background technology

Along with the aggravation of China's aging, the health care of elderly population causes the concern of society.Hemiplegia is a kind of disease common in elderly population, and inconvenient walking, gait gently then can be caused to be out of shape, heavy then be unable to leave the bed, completely lose viability.If can not obtain effective rehabilitation, their lower extremity motor function may cannot recover forever, has had a strong impact on to their normal life.Healing robot is the new opplication that robot and rehabilitation medical combine, and healing robot, in conjunction with different robot control methods, can provide initiatively and passive rehabilitation training for lower limb paralysis patient.Under the increasing background of China's aging population, the application of healing robot not only can alleviate the hard work of rehabilitation physical therapy teacher in Traditional Rehabilitation treatment, the more important thing is that healing robot can provide rehabilitation whenever and wherever possible according to the wish of patient, therapeutic process can be monitored and record in real time, and then can update rehabilitation after analyzing these data, so healing robot can provide scientificlly and effectively rehabilitation.

Chinese invention patent 201110404114.9 discloses a kind of passive self-adaptation control method of master of upper and lower limbs rehabilitation training robot, but its so-called main Passive Control is the simple deviation according to motor power voltage value judges whether patient has initiatively intention, but but do not provide this processing procedure according to and this method whether there is reliability.So-called adaptive control is also one and simply judges, takes the initiative or passive control methods in certain scope according to voltage error value.This adaptive control can not overcome robot Parameters variation or disturbance to the impact of robot control system.

Chinese invention patent 201010561379.5 discloses a kind of motion control method of lower limb rehabilitative robot, for two stages in the rehabilitation course of patient have planned two kinds of training modes in this invention, i.e. and active and passive exercise pattern.Judge the motion intention of patient according to human computer interaction's power, then produce gait track by adaptive controller, carry out active training.This invention gives robot system two close cycles control method, does not relate to the design of specific algorithm, and this control method does not consider that robot system external disturbance and Parameters variation are on the impact of the control of robot.

Analyze existing invention can find, control method in existing invention focuses mostly in the design to robot system top level control method, such as, active and passive exercise, but seldom relate to and realizing initiatively and the design of the bottom control method of passive exercise, such as, consider when realizing main Passive Control, how control machine person joint realizes movement locus accurately.

Summary of the invention

The object of the invention is to overcome the deficiencies in the prior art, a kind of control method of lower limb exoskeleton rehabilitation robot rehabilitation training is provided, by the mode of bottom control, control lower limb rehabilitation robot and do accurate motion in accordance with the instructions.

For achieving the above object, the control method of a kind of lower limb exoskeleton rehabilitation robot rehabilitation training of the present invention, is characterized in that, comprise the following steps:

(1), to lower limb exoskeleton rehabilitation robot carry out modeling, the governing equation obtaining lower limb exoskeleton rehabilitation robot is:

$M\left(q\left(t\right)\right)\stackrel{\·\·}{q}\left(t\right)+V(q\left(t\right),\stackrel{\·}{q}\left(t\right))+G\left(q\left(t\right)\right)=\mathrm{\τ}\left(t\right)---\left(1\right)$

Wherein, q (t) represents the angle of rotation of knee joint and hip joint; represent the velocity of rotation of knee joint and hip joint; represent the rotation acceleration of knee joint and hip joint; τ (t) is adaptive controller; M (q (t)) represents the inertial matrix of lower limb exoskeleton rehabilitation robot model; G (q (t)) represents the gravity of lower limb exoskeleton rehabilitation robot model; represent the centrifugal and coriolis force item matrix of lower limb exoskeleton robot model;

Linearization is carried out to formula (1), obtains:

Wherein, for the matrix of function of time composition, and with angle of rotation q (t), the velocity of rotation of knee joint and hip joint relevant; represent the unknown parameter in robot model;

(2), adaptive controller τ (t) is built, and adaptive control rate

Wherein, be respectively: the estimated value of inertial matrix, centrifugal and the estimated value of coriolis force item matrix, the estimated value of gravity;

for estimation, As time goes on, constantly level off to true value for first order derivative;

E (t)=q _dt ()-q (t), represents tracking error vector, $\stackrel{\·}{e}\left(t\right)=d\left(e\left(t\right)\right)/\mathrm{dt},\stackrel{\·\·}{e}\left(t\right)=d^2\left(e\left(t\right)\right)/{\mathrm{dt}}^2,$ Wherein, q _dt angle of rotation that () arrives for knee joint and hip joint expection;

$\mathrm{\Ω}\left(t\right)={[\mathrm{0,0},\stackrel{\·}{e}{\left(t\right)}^T]}^T;$

K _vand K _pit is ride gain parameter, obtained by training in advance, training method is: in advance by MATLAB software emulation lower limb exoskeleton rehabilitation robot system, other parameters in boarder controller τ (t) are set, simulation training is carried out again according to the governing equation in boarder controller τ (t) and step (1), its training objective is tracking error e (the t)≤M making each joint in lower limb exoskeleton rehabilitation robot system, 0≤M < 100;

(3), at moment t, photoelectric encoder is adopted to gather angle of rotation q (t) and the velocity of rotation of knee joint and hip joint

(4) actual parameter, step (3) obtained substitutes into step (2) and obtains adaptive control rate thus obtaining adaptive controller τ (t) of moment t, drive unit applies acting force according to adaptive controller τ (t) to lower limb rehabilitation robot again, and driving joint turns to q _d(t).

Goal of the invention of the present invention is achieved in that

The control method of lower limb exoskeleton rehabilitation robot rehabilitation training of the present invention, by carrying out modeling to lower limb exoskeleton rehabilitation robot, obtain model equation, build adaptive controller and control rate computing formula again, for wherein undetermined ride gain parameter, obtained by training in advance, then actual angle of rotation and the velocity of rotation recording lower limb exoskeleton rehabilitation robot knee joint and hip joint, finally actual parameter is substituted into adaptive control rate computing formula, thus obtain the adaptive controller of moment t, drive unit applies acting force according to adaptive control rate to lower limb exoskeleton rehabilitation robot.

Meanwhile, the control method of a kind of lower limb exoskeleton rehabilitation robot of the present invention rehabilitation training also has following beneficial effect:

(1), present invention employs adaptive control algorithm, be applied to lower limb exoskeleton rehabilitation robot system, for lower limb paralysis, patient provides rehabilitation training, even if occurring under disturbance and inherent parameters change, can solve the not enough problem of control accuracy that robot model's Parameters variation brings equally;

(2), the present invention compared to Traditional Rehabilitation training patterns, advantage is patient by the constraint in time, place, and training process can record, assess;

(3), the present invention is when designing adaptive controller, add trace vector, namely, when lower limb exoskeleton rehabilitation robot is by driving instruction motion, can judge whether that controlling lower limb exoskeleton rehabilitation robot carries out rehabilitation exercise by driving instruction by the mode of following the tracks of.

Accompanying drawing explanation

Fig. 1 Fig. 1 is the physical construction schematic diagram of lower limb exoskeleton rehabilitation robot;

Fig. 2 is the control method process flow diagram of lower limb exoskeleton rehabilitation robot rehabilitation training of the present invention;

Fig. 3 is the control method theory diagram of lower limb exoskeleton rehabilitation robot rehabilitation training of the present invention;

Fig. 4 is that kneed rotational angle follows the tracks of schematic diagram;

Fig. 5 is that the rotational angle of hip joint follows the tracks of schematic diagram;

Fig. 6 is the rotational angle tracking error schematic diagram of knee joint and hip joint.

Embodiment

Below in conjunction with accompanying drawing, the specific embodiment of the present invention is described, so that those skilled in the art understands the present invention better.Requiring particular attention is that, in the following description, when perhaps the detailed description of known function and design can desalinate main contents of the present invention, these are described in and will be left in the basket here.

Embodiment

Fig. 1 is the physical construction schematic diagram of lower limb exoskeleton rehabilitation robot.

In the present embodiment, as shown in Figure 1, every bar leg of lower limb exoskeleton rehabilitation robot all has 2 degree of freedom, and namely hip joint and knee joint respectively have one degree of freedom.In adaptive controller design process below, be all design based on the 2DOF model of single leg.

Fig. 2 is the control method process flow diagram of lower limb exoskeleton rehabilitation robot rehabilitation training of the present invention.

As shown in Figure 2, in the present embodiment, a kind of control method of lower limb exoskeleton rehabilitation robot rehabilitation training, comprises the following steps:

S201, carry out modeling to lower limb exoskeleton rehabilitation robot, the model equation obtaining lower limb exoskeleton rehabilitation robot is:

$M\left(q\left(t\right)\right)\stackrel{\&CenterDot;\&CenterDot;}{q}\left(t\right)+V(q\left(t\right),\stackrel{\&CenterDot;}{q}\left(t\right))+G\left(q\left(t\right)\right)=\mathrm{\&tau;}\left(t\right)---\left(1\right)$

Wherein, q (t)=[q ₁(t), q ₂(t)] ^t, represent knee joint q ₁(t) and hip joint q ₂the angle of rotation of (t); represent the velocity of rotation of knee joint and hip joint; represent the rotation acceleration of knee joint and hip joint; τ (t) is adaptive controller; M (q (t)) represents the inertial matrix of lower limb exoskeleton rehabilitation robot model; G (q (t)) represents the gravity of lower limb exoskeleton rehabilitation robot model; represent the centrifugal and coriolis force item matrix of lower limb exoskeleton robot model;

Wherein,

$M\left(q\left(t\right)\right)=\left[\begin{array}{cc}(\frac{1}{4}{m}_1+m_2)a_1^2+\frac{1}{4}{m}_2{a}_2^2+m_2{a}_1{a}_2\cos \left(q_2\left(t\right)\right)& \frac{1}{4}{m}_2{a}_2^2+m_2{a}_1{a}_2\cos \left(q_2\left(t\right)\right)\\ \frac{1}{4}{m}_2{a}_2^2+m_2{a}_1{a}_2\cos \left(q_2\left(t\right)\right)& \frac{1}{4}{m}_2{a}_2^2\end{array}\right]$

$V(q\left(t\right),\stackrel{\&CenterDot;}{q}\left(t\right))=\frac{1}{2}{m}_2{a}_1{a}_2\sin \left(q_2\left(t\right)\right)\left[\begin{array}{cc}-2{\stackrel{\&CenterDot;}{q}}_2\left(t\right)& -{\stackrel{\&CenterDot;}{q}}_2\left(t\right)\\ {\stackrel{\&CenterDot;}{q}}_1\left(t\right)& 0\end{array}\right]$

$G\left(q\left(t\right)\right)=\left[\begin{array}{c}(\frac{1}{2}{m}_1+m_2){\mathrm{ga}}_1\cos \left(q_1\left(t\right)\right)+\frac{1}{2}{m}_2{\mathrm{ga}}_2\cos (q_1\left(t\right)+q_2\left(t\right))\\ \frac{1}{2}{m}_2{\mathrm{ga}}_2\cos (q_1\left(t\right)+q_2\left(t\right))\end{array}\right]$

Wherein: m ₁, m ₂represent the quality of knee joint and hip joint respectively; a ₁, a ₂represent the length of knee joint and hip joint respectively, g is gravity constant;

Order: then obtain:

Linearization is carried out to formula (1), obtains:

Wherein, for the matrix of function of time composition, and with angle of rotation q (t), the velocity of rotation of knee joint and hip joint relevant; represent the unknown parameter in robot model;

After the model equation foundation of lower limb exoskeleton rehabilitation robot, need the correlation parameter determined in equation, and incorporating parametric controls the motion of lower limb exoskeleton rehabilitation robot, below in step, being but described in detail to correlation parameter, as follows:

S202, structure adaptive controller τ (t), and adaptive control rate

Wherein, be respectively: the estimated value of inertial matrix, centrifugal and the estimated value of coriolis force item matrix, the estimated value of gravity;

for estimation, As time goes on, constantly level off to true value for first order derivative;

E (t)=q _dt ()-q (t), represents tracking error vector, wherein, q _dt angle of rotation that () arrives for knee joint and hip joint expection;

$\mathrm{\&Omega;}\left(t\right){=[\mathrm{0,0},\stackrel{\&CenterDot;}{e}{\left(t\right)}^T]}^T;$

K _vand K _pit is ride gain parameter, obtained by training in advance, training method is: in advance by MATLAB software emulation lower limb exoskeleton rehabilitation robot system, other parameters in boarder controller τ (t) are set, simulation training is carried out again according to the governing equation in boarder controller τ (t) and step S201, its training objective is tracking error e (the t)≤M making each joint in lower limb exoskeleton rehabilitation robot system, 0≤M < 100; In the present embodiment as M=5, be met the K of training objective _vand K _p, be ride gain parameter required for the present invention;

S203, at moment t, photoelectric encoder is adopted to gather angle of rotation q (t) and the velocity of rotation of knee joint and hip joint

S204, the actual parameter obtained by step S203 substitute into step S202 and obtain adaptive control rate thus obtaining adaptive controller τ (t) of moment t, drive unit applies acting force according to adaptive controller τ (t) to lower limb rehabilitation robot again, and driving joint turns to q _d(t).

For practicality of the present invention is described, stability checking is carried out to lower limb rehabilitation robot system below.

Definition tracking error is: e (t)=q _d(t)-q (t);

By t unknown parameter estimated value be updated to formula (2), can obtain:

Can obtain according to formula (1), (2):

Can obtain according to formula (3) and (6) again:

Definition unknown parameter error be: therefore, can obtain according to error definition:

Formula (8) is rewritten into following form:

Single order and second derivative are asked to tracking error e (t), obtain:

$\begin{array}{c}\stackrel{\&CenterDot;}{e}\left(t\right)={\stackrel{\&CenterDot;}{q}}_d\left(t\right)-\stackrel{\&CenterDot;}{q}\left(t\right)\\ \stackrel{\&CenterDot;\&CenterDot;}{e}\left(t\right)={\stackrel{\&CenterDot;\&CenterDot;}{q}}_d\left(t\right)-\stackrel{\&CenterDot;\&CenterDot;}{q}\left(t\right)\end{array}---\left(11\right)$

Order $E\left(t\right)={[e\left(t\right),\stackrel{\&CenterDot;}{e}\left(t\right)]}^T;A=\left[\begin{array}{cc}{O}_n& I_n\\ K_p& K_v\end{array}\right],B=\left[\begin{array}{c}{O}_n\\ O_n\end{array}\right],$ O _nbe 2 × 1 rank 0 vectors, I _nbe 2 × 1 rank vector of unit length;

Then formula (10) can be write as vector form:

Choosing Liapunov function is:

Wherein, P is 4 × 4 symmetric positive definite scalar matrixes; Γ is 5 × 5 diagonal angle positive definite scalar matrixes.

First order derivative is asked to formula (13), obtains:

According to symmetry, and by (12) substitute into (14) can obtain:

Wherein, Q is positive definite symmetric matrices, and meets Li Yapuduofu equation

A ^TP+PA＝-Q (16)

Adaptive rate in formula (4) is equivalent to

Getting Γ, P is unit diagonal matrix, and (17) substitution (15) are obtained:

$\stackrel{\&CenterDot;}{V}\left(t\right)=-E^T\left(t\right)\mathrm{QE}\left(t\right)\&le;0---\left(18\right)$

Known according to formula (18), Lyapunov function V (t) lower bound is zero; Known according to formula (13), E (t), all bounded; According to the definition of tracking error, bounded, meanwhile, also be bounded.Known according to formula (12) again bounded; Therefore, by Rayleigh-inner hereby theorem can obtain, lower limb exoskeleton rehabilitation robot system is asymptotically stability.

Fig. 3 is the control method theory diagram of lower limb exoskeleton rehabilitation robot rehabilitation training of the present invention.

In the present embodiment, as shown in Figure 3, adaptive controller, according to expecting that rehabilitation track and the actual path data fed back calculate controlling torque, then exports corresponding torque by motor, and then control motion, and then provide lower limb rehabilitation training for paralytic.

Fig. 4 is that kneed rotational angle follows the tracks of schematic diagram.

Fig. 5 is that the rotational angle of hip joint follows the tracks of schematic diagram.

Under the effect of adaptive controller τ (t), as shown in Figure 4, the knee joint actual rotation angle q of lower limb exoskeleton rehabilitation robot ₁t () just achieves for expected angle q in 1s _1dt () high-precision position is followed the tracks of, and tracking error is less than 1 after t=1s; As shown in Figure 5, the hip joint actual rotation angle q of lower limb exoskeleton rehabilitation robot ₂t () just achieves for expected angle q in 1.1s _1dt () high-precision position is followed the tracks of, after t=1.1s, tracking error is less than 1 equally.

Fig. 6 is the rotational angle tracking error schematic diagram of knee joint and hip joint.

Under the effect of adaptive controller, after t=1s, the tracking error e of knee joint and hip joint ₁(t), e ₂t () is all less than 1, meet setting index, complete the control objectives that position is followed the tracks of, demonstrate the validity of controller.And e ₁(t), e ₂t () is tending towards 0 after 2s, the controller which illustrating design has good control performance, can reach effect needed for us satisfactorily.Thus, achieve the accuracy that fast driving system arrives precalculated position control motion.

Although be described the illustrative embodiment of the present invention above; so that those skilled in the art understand the present invention; but should be clear; the invention is not restricted to the scope of embodiment; to those skilled in the art; as long as various change to limit and in the spirit and scope of the present invention determined, these changes are apparent, and all innovation and creation utilizing the present invention to conceive are all at the row of protection in appended claim.

Claims (1)

1. a control method for lower limb exoskeleton rehabilitation robot rehabilitation training, is characterized in that, comprises the following steps:

(1), to lower limb exoskeleton rehabilitation robot carry out modeling, the governing equation obtaining lower limb exoskeleton rehabilitation robot is:

$M\left(q\left(t\right)\right)\stackrel{\&CenterDot;\&CenterDot;}{q}\left(t\right)+V(q\left(t\right),\stackrel{\&CenterDot;}{q}\left(t\right))+G\left(q\left(t\right)\right)=\mathrm{\&tau;}\left(t\right)---\left(1\right)$

Wherein, q (t) represents the angle of rotation of knee joint and hip joint; represent the velocity of rotation of knee joint and hip joint; represent the rotation acceleration of knee joint and hip joint; τ (t) is adaptive controller; M (q (t)) represents the inertial matrix of lower limb exoskeleton rehabilitation robot model; G (q (t)) represents the gravity of lower limb exoskeleton rehabilitation robot model; represent the centrifugal and coriolis force item matrix of lower limb exoskeleton robot model;

Linearization is carried out to formula (1), obtains:

Wherein, for the matrix of function of time composition, and with angle of rotation q (t), the velocity of rotation of knee joint and hip joint relevant; represent the unknown parameter in robot model;

(2), adaptive controller τ (t) is built, and adaptive control rate

Wherein, be respectively: the estimated value of inertial matrix, the estimated value of coriolis force item matrix, the estimated value of gravity;

for estimation, As time goes on, constantly level off to true value for first order derivative;

E (t)=q _dt ()-q (t), represents tracking error vector, $\stackrel{\&CenterDot;}{e}\left(t\right)=d\left(e\left(t\right)\right)/\mathrm{dt},\stackrel{\&CenterDot;\&CenterDot;}{e}\left(t\right)=d^2\left(e\left(t\right)\right)/d{t}^2,$ Wherein, q _dt angle of rotation that () arrives for knee joint and hip joint expection;

$\mathrm{\&Omega;}\left(t\right)={[\mathrm{0,0},\stackrel{\&CenterDot;}{e}{\left(t\right)}^T]}^T;$

K _vand K _pit is ride gain parameter, obtained by training in advance, training method is: in advance by MATLAB software emulation lower limb exoskeleton rehabilitation robot system, other parameters in boarder controller τ (t) are set, simulation training is carried out again according to the governing equation in boarder controller τ (t) and step (1), its training objective is tracking error e (the t)≤M% making each joint in lower limb exoskeleton rehabilitation robot system, 0≤M < 100;

(3), at moment t, photoelectric encoder is adopted to gather angle of rotation q (t) and the velocity of rotation of knee joint and hip joint

(4) actual parameter, step (3) obtained substitutes into step (2) and obtains adaptive control rate thus obtaining adaptive controller τ (t) of moment t, drive unit applies acting force according to adaptive controller τ (t) to lower limb rehabilitation robot again, and driving joint turns to q _d(t).