CN104252173A - Walking control method of biped walking robot - Google Patents

Abstract

The invention discloses a walking control method of a biped walking robot. The method includes the steps: firstly, setting walking parameters of the biped walking robot; secondly, computing the center of mass and a reference (zero moment point) track of the biped walking robot; thirdly, computing a working pattern of the biped walking robot by means of inverse kinematics; fourthly, computing the ground friction coefficient needed for non-slip in walking of the biped walking robot, and judging whether the biped walking robot slips in walking or not; fifthly, if so, regulating the walking parameters of the biped walking robot, and repeating the second, third and fourth steps. By the method, whether the biped walking robot slips on the ground with low friction coefficient or not can be judged, the stepping parameters of the walking pattern of the biped walking robot are regulated to reduce the minimum ground friction coefficient needed by the walking pattern, and walking capability of the biped walking robot on the ground with the low friction coefficient is improved.

Description

The walking control method of bipod walking robot

Technical field

The present invention relates to a kind of control method of robot, specifically, relate to a kind of walking control method of robot.

Background technology

The stability control method of current bipod walking robot is all adopt ZMP stability criterion, ZMP refers to that the level of torque component of ground force is the application point of zero, ZMP stability criterion refers to that the precondition of bipod walking robot walking stability is that ZMP is arranged in support polygon all the time, and support polygon refers to the pin of biped robot and the convex set of ground-engaging area.Meet the physical significance directly perceived of the double feet walking of ZMP stability criterion, refer in biped robot's gait processes and can not topple over.The gait stability control of existing biped robot is all each joint motions planning bipod walking robot based on ZMP stability criterion, is positioned at support polygon all the time to meet the ZMP of bipod walking robot in gait processes.

But existing ZMP stability criterion has certain limitation, the well-known bipod walking robot expert Kajita of Japanese AIST once pointed out that ZMP had three kinds of situations to apply, and one of them is exactly the situation occurring with ground at the bottom of robot foot to skid.In fact, ZMP stability criterion, to ignore premised on frictional ground force, is namely thought that the friction force between biped robot's pin and ground is enough large, the situation of skidding can not be occurred between the pin of bipod walking robot and ground.This prerequisite is feasible under ecotopia, and naturally, the walking movement based on the bipod walking robot of ZMP stability criterion planning meets the stability requirement of walking.But for the situation that the friction force between the pin of bipod walking robot and ground is less, based on the walking movement of the bipod walking robot of ZMP stability criterion planning, can only ensure can not topple in robot gait processes, but cannot ensure that robot can not skid.

Summary of the invention

The object of the present invention is to provide a kind of walking control method of bipod walking robot, can judge whether bipod walking robot can skid on low-friction coefficient ground, simultaneously by regulating the walking parameter of bipod walking robot Walking Mode, reduce the minimum ground friction factor needed for Walking Mode, improve the walking ability of bipod walking robot on low-friction coefficient ground.

To achieve these goals, the technical solution adopted in the present invention is as follows:

A walking control method for bipod walking robot, comprises the following steps: step one: the walking parameter arranging described bipod walking robot; Step 2: calculate the barycenter of described bipod walking robot and the track with reference to ZMP; Step 3: the Walking Mode being gone out described bipod walking robot by the computation of inverse-kinematics; Step 4: calculate the non-slip required ground friction coefficient of described bipod walking robot walking, judge whether described bipod walking robot walking skids; Step 5: if described bipod walking robot walking is skidded, then regulate the described walking parameter of described bipod walking robot, repeats above-mentioned steps two, step 3 and step 4.

Further, the non-slip required described ground friction coefficient of described bipod walking robot walking is calculated according to the horizontal force component of the intermolecular forces on described bipod walking robot feet and ground and vertical force component in step 4.

Further, the non-slip required described ground friction coefficient of the described bipod walking robot walking calculated in step 4 is compared with actual ground friction factor, if required described ground friction coefficient is less than described actual ground friction factor, then described bipod walking robot walking is non-slip, otherwise then skids.

Further, if described bipod walking robot walking is non-slip in step 4, then the described Walking Mode of planned described bipod walking robot meets ZMP stability requirement.

Further, the walking model that the generation of the described Walking Mode of bipod walking robot described in step 3 is is bipod walking robot with desk-dolly.

Further, the generation of the described Walking Mode of bipod walking robot described in step 3 is for system inputs with the differential of the acceleration of barycenter.

Further, the generation of the described Walking Mode of bipod walking robot described in step 3 is that the parameter in evaluation function is for setting described walking parameter by the minimized method tracing control of evaluation function with reference to ZMP track.

Further, the parameter in described evaluation function is for setting walking speed and the step-length of described walking parameter.

Further, the described Walking Mode of the described bipod walking robot generated in step 3 meets the ZMP stability criterion of double feet walking, combines with stability criterion of whether skidding.

Compared with prior art, the walking control method of bipod walking robot of the present invention, can judge whether bipod walking robot can skid on low-friction coefficient ground, simultaneously by regulating the walking parameter of bipod walking robot Walking Mode, reduce the minimum ground friction factor needed for Walking Mode, improve the walking ability of bipod walking robot on low-friction coefficient ground.

Accompanying drawing explanation

Fig. 1 is the schematic diagram of the bipod walking robot of the walking control method of bipod walking robot of the present invention;

Fig. 2 is the schematic diagram of the desk-little vehicle model of the walking control method of bipod walking robot of the present invention;

Fig. 3 is the schematic flow sheet of the walking control method of bipod walking robot of the present invention.

Embodiment

Be described further below in conjunction with the walking control method of the drawings and specific embodiments to bipod walking robot of the present invention.

The invention provides the method for a kind of bipod walking robot 1 in the walking of low-friction coefficient ground, by designing a kind of walking speed and the adjustable Walking Mode generation method of step-length, and take the stability criterion of sliding friction force constraint into consideration, regulate the parameter of Walking Mode, setting is applicable to the Walking Mode of ground friction factor, improves the walking control method of bipod walking robot 1.

Refer to Fig. 1, the decision method of described bipod walking robot 1 walking stability during consideration sliding friction force constraint, specific as follows:

Described bipod walking robot 1 on the ground walking time, right crus of diaphragm 11R(is feet) and ground between acting force be f, its horizontal component is f _t, vertical component is f _z, the friction factor on ground is μ, does not slide and need meet the following conditions between described right crus of diaphragm 11R and ground

f _t<μf _z (1)

For known Walking Mode, the momentum of described bipod walking robot 1 can be expressed as

$p=\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{m}_i{\stackrel{\&CenterDot;}{c}}_i---\left(2\right)$

Wherein m _iand c _ibe respectively quality and the centroid position of each connecting rod.

The described horizontal component f of ground force f can be obtained to formula (2) differential _twith described vertical component f _zfor

$f_t=\sqrt{{\stackrel{\&CenterDot;}{p}}_x^2+{\stackrel{\&CenterDot;}{p}}_y^2}$

(3)

$f_z={\stackrel{\&CenterDot;}{p}}_z+\mathrm{Mg}$

Wherein M is the overall weight of described bipod walking robot 1, and g is acceleration of gravity.

According to formula (3) can obtain the walking of described bipod walking robot 1 do not occur slide ground friction coefficient be

${\mathrm{\&mu;}}_{\mathrm{nec}}=\frac{f_t}{f_z}---\left(4\right)$

Work as μ _necduring≤μ, can not slide on the ground during described bipod walking robot 1 walking, coordinate ZMP stability criterion, namely when in described bipod walking robot 1 gait processes, the level of torque component of described ground force f is that the point of zero is in the support polygon that described bipod walking robot 1 feet and ground are formed, described bipod walking robot 1 can not be toppled over, and can realize the proper walking of described bipod walking robot 1 and stablize.The method as the expansion of existing ZMP stability criterion, can make more strict judgement to the stability of described bipod walking robot 1, makes the range of application of described bipod walking robot 1 wider.

The present invention is based on a kind of Walking Mode generation method that above-mentioned new stability criterion proposes adjustable walking speed and step-length, to adapt to the walking situation of described bipod walking robot 1 in the less situation of ground friction coefficient, specific as follows:

Refer to Fig. 1 and Fig. 2, described bipod walking robot 1, when the walking of low-friction coefficient ground, can adopt classical desk-dolly model representation:

$p_x=x-\frac{z_c}{g}\stackrel{\&CenterDot;\&CenterDot;}{x}$

(5)

$p_y=y-\frac{z_c}{g}\stackrel{\&CenterDot;\&CenterDot;}{y}$

Wherein x and y represents the projection of described bipod walking robot 1 barycenter in level ground, p _xand p _yrepresent the position of ZMP.

The differential supposing described bipod walking robot 1 barycenter acceleration is the input variable of system, namely

${\mathrm{in}}_x=\stackrel{\&CenterDot;\&CenterDot;\&CenterDot;}{x}$

(6)

${\mathrm{in}}_y=\stackrel{\&CenterDot;\&CenterDot;\&CenterDot;}{y}$

Then

$\frac{d}{\mathrm{dt}}\left[\begin{array}{c}x\\ \stackrel{\&CenterDot;}{x}\\ \stackrel{\&CenterDot;\&CenterDot;}{x}\end{array}\right]=\left[\begin{array}{ccc}0& 1& 0\\ 0& 0& 1\\ 0& 0& 0\end{array}\right]\left[\begin{array}{c}x\\ \stackrel{\&CenterDot;}{x}\\ \stackrel{\&CenterDot;\&CenterDot;}{x}\end{array}\right]+\left[\begin{array}{c}0\\ 0\\ 1\end{array}\right]u_z$

$\frac{d}{\mathrm{dt}}\left[\begin{array}{c}y\\ \stackrel{\&CenterDot;}{y}\\ \stackrel{\&CenterDot;\&CenterDot;}{y}\end{array}\right]=\left[\begin{array}{ccc}0& 1& 0\\ 0& 0& 1\\ 0& 0& 0\end{array}\right]\left[\begin{array}{c}y\\ \stackrel{\&CenterDot;}{y}\\ \stackrel{\&CenterDot;\&CenterDot;}{y}\end{array}\right]+\left[\begin{array}{c}0\\ 0\\ 1\end{array}\right]u_y---\left(7\right)$

Meanwhile, formula (5) can be expressed as

$p_x=\left[\begin{array}{ccc}1& 0& -\frac{z_c}{g}\end{array}\right]\left[\begin{array}{c}x\\ \stackrel{\&CenterDot;}{x}\\ \stackrel{\&CenterDot;\&CenterDot;}{x}\end{array}\right]$

(8)

$p_y=\left[\begin{array}{ccc}1& 0& -\frac{z_c}{g}\end{array}\right]\left[\begin{array}{c}y\\ \stackrel{\&CenterDot;}{y}\\ \stackrel{\&CenterDot;\&CenterDot;}{y}\end{array}\right]$

Convolution (7) and formula (8), carrying out discretize to it can obtain

$\left\{\begin{array}{c}{x}_{k+1}={\mathrm{Ax}}_k+{\mathrm{bu}}_{x_k}\\ y_{k+1}={\mathrm{Ay}}_k+{\mathrm{bu}}_{y_k}\\ p_{x_k}={\mathrm{cx}}_k\\ p_{y_k}={\mathrm{cy}}_k\end{array}\right.---\left(9\right)$

Wherein

$x_k={\left[\begin{array}{ccc}x\left(\mathrm{k\&Delta;t}\right)& \stackrel{\&CenterDot;}{x}\left(\mathrm{k\&Delta;t}\right)& \stackrel{\&CenterDot;\&CenterDot;}{x}\left(\mathrm{k\&Delta;t}\right)\end{array}\right]}^T$

$y_k={\left[\begin{array}{ccc}y\left(\mathrm{k\&Delta;t}\right)& \stackrel{\&CenterDot;}{y}\left(\mathrm{k\&Delta;t}\right)& \stackrel{\&CenterDot;\&CenterDot;}{y}\left(\mathrm{k\&Delta;t}\right)\end{array}\right]}^T$

u _xk=u _x(kΔt)

u _yk=u _y(kΔt)

p _xk=p _x(kΔt)

p _yk=p _y(kΔt)

$A=\left[\begin{array}{ccc}1& \mathrm{\&Delta;t}& {\mathrm{\&Delta;t}}^2/2\\ 0& 1& \mathrm{\&Delta;t}\\ 0& 0& 1\end{array}\right]$

$b=\left[\begin{array}{c}{\mathrm{\&Delta;t}}^3/6\\ {\mathrm{\&Delta;t}}^2/2\\ 1\end{array}\right]---\left(10\right)$

$c=\left[\begin{array}{ccc}1& 0& -z_c/g\end{array}\right]$

In order to make the p of planning _xkand p _yktrack reference ZMP as much as possible with here realize tracing control by minimizing of evaluation function, evaluation function is as follows

$J=\underset{j=1}{\overset{\&infin;}{\mathrm{\&Sigma;}}}\{Q_x{(p_{x_j}^{\mathrm{ref}}-p_{x_j})}^2+R_x{u}_{x_j}\}+\underset{j=1}{\overset{\&infin;}{\mathrm{\&Sigma;}}}\{Q_y{(p_{y_j}^{\mathrm{ref}}-p_{y_j})}^2+R_y{u}_{y_j}\}---\left(11\right)$

Q in formula _x, Q _y, R _x, R _yfor positive weighting coefficient, wherein

$Q_x=\left[\begin{array}{ccc}{Q}_{x_1}& 0& 0\\ 0& Q_{x_2}& 0\\ 0& 0& Q_{x_3}\end{array}\right],Q_y=\left[\begin{array}{ccc}{Q}_{y_1}& 0& 0\\ 0& Q_{y_2}& 0\\ 0& 0& Q_{y_3}\end{array}\right]---\left(12\right)$

Wherein with the coefficient relevant with walking speed and step-length, namely different by setting with value can regulate speed and the step-length of walking, and the now input of system can obtain according to preview control theory

$u_{x_k}=-{\mathrm{Kx}}_k+\left[\begin{array}{cccc}{f}_{x_1}& f_{x_2}& \&CenterDot;\&CenterDot;\&CenterDot;& f_{x_N}\end{array}\right]\left[\begin{array}{c}{p}_{x_{k+1}}^{\mathrm{ref}}\\ p_{x_{k+2}}^{\mathrm{ref}}\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ p_{x_{k+N}}^{\mathrm{ref}}\end{array}\right]$

(13)

$u_{y_k}=-{\mathrm{Ky}}_k+\left[\begin{array}{cccc}{f}_{y_1}& f_{y_2}& \&CenterDot;\&CenterDot;\&CenterDot;& f_{y_N}\end{array}\right]\left[\begin{array}{c}{p}_{y_{k+1}}^{\mathrm{ref}}\\ p_{y_{k+2}}^{\mathrm{ref}}\\ \&CenterDot;\\ \&CenterDot;\\ \&CenterDot;\\ p_{y_{k+N}}^{\mathrm{ref}}\end{array}\right]$

In formula

K=(R+b ^TPb) ^-1b ^TPA (14)

f _i=(R+b ^TPb) ^-1b ^T(A-bK) ^T×(i-1)c ^TQ

Wherein matrix P is non trivial solution below

P=A ^TPA+c ^TQc-A ^TPb(R+b ^TPb) ^-1b ^TPA (15)

So far the Walking Mode of described bipod walking robot 1 can be calculated, the i.e. track of described bipod walking robot 1 barycenter and the track with reference to ZMP, to be based upon the coordinate system ∑ 0 of the waist of described bipod walking robot 1 for basis coordinates system, with ∑ w for world coordinate system, as shown in Figure 1.Walking Mode (the p of described bipod walking robot 1 can be calculated by inverse kinematics ₀, R ₀, θ), wherein (p ₀, R ₀) be the position of base coordinate system ∑ 0 in world coordinate system ∑ w and attitude, θ is the angle in described each joint of bipod walking robot 1.

In sum, refer to Fig. 1 and Fig. 3, the walking control method of described bipod walking robot 1 disclosed by the invention, comprises the following steps:

Step one S1: the related coefficient that the walking parameter of bipod walking robot 1 described in formula (11) and formula (12) is set.

Step 2 S3: calculate the barycenter of described bipod walking robot 1 and the track with reference to ZMP according to formula (13).

Step 3 S5: be basis coordinates system with ∑ 0, ∑ w is world coordinate system, is gone out the Walking Mode (p of described bipod walking robot 1 by the computation of inverse-kinematics ₀, R ₀, θ).

The generation of described Walking Mode take desk-dolly as the walking model of described bipod walking robot 1, with the differential of the acceleration of barycenter for system inputs, by the minimized method tracing control of evaluation function with reference to ZMP track, the parameter in evaluation function is for setting the walking parameter such as walking speed and step-length.The described Walking Mode of the described bipod walking robot 1 generated meets the ZMP stability criterion of double feet walking, combine with stability criterion of whether skidding, form more strict described bipod walking robot 1 walking stability criterion, widen the walking movement of described bipod walking robot 1 in imperfect environment.

Step 4 S7: calculate the non-slip required ground friction coefficient μ of described bipod walking robot 1 walking according to formula (2), formula (4) and formula (7) _nec, judge whether the walking of described bipod walking robot 1 skids.

Whether described bipod walking robot 1 skids in ground walking, is the horizontal force component f of the intermolecular forces f according to described double feet walking machine 1 feet and ground _twith vertical force component f _zcalculate the non-slip required minimum ground coefficientoffrictionμ of described bipod walking robot 1 walking _nec, and compared with actual ground coefficientoffrictionμ, if required minimum ground coefficientoffrictionμ _necbe less than ground friction coefficient μ, then described bipod walking robot 1 is non-slip, otherwise then skids.If the walking of described bipod walking robot 1 is non-slip, then the described Walking Mode of planned described bipod walking robot 1 meets ZMP stability requirement.

Step 5 S9: if described bipod walking robot 1 walking is skidded, then regulate the described walking parameter of described bipod walking robot 1, repeats above-mentioned steps two, step 3 and step 4.

The walking control method of bipod walking robot 1 of the present invention, can judge whether described bipod walking robot 1 can skid on low-friction coefficient ground, simultaneously by regulating the walking parameter of described bipod walking robot 1 Walking Mode, reduce the minimum ground friction factor needed for Walking Mode, improve the walking ability of described bipod walking robot 1 on low-friction coefficient ground.

More than describe preferred embodiment of the present invention in detail, should be appreciated that those of ordinary skill in the art just design according to the present invention can make many modifications and variations without the need to creative work.Therefore, all technician in the art according to the present invention's design on prior art basis by logic analysis, reasoning or according to the available technical scheme of limited experiment, all should by among the determined protection domain of these claims.

Claims (9)

1. a walking control method for bipod walking robot, is characterized in that, comprises the following steps:

Step one: the walking parameter that described bipod walking robot is set;

Step 2: calculate the barycenter of described bipod walking robot and the track with reference to ZMP;

Step 3: the Walking Mode being gone out described bipod walking robot by the computation of inverse-kinematics;

Step 4: calculate the non-slip required ground friction coefficient of described bipod walking robot walking, judge whether described bipod walking robot walking skids;

Step 5: if described bipod walking robot walking is skidded, then regulate the described walking parameter of described bipod walking robot, repeats above-mentioned steps two, step 3 and step 4.

2. the walking control method of bipod walking robot as claimed in claim 1, is characterized in that: in step 4 according to the horizontal force component of described bipod walking robot feet and the intermolecular forces on ground and vertical force component calculate described bipod walking robot walking non-slip needed for described ground friction coefficient.

3. the walking control method of bipod walking robot as claimed in claim 2, it is characterized in that: the non-slip required described ground friction coefficient of the described bipod walking robot walking calculated in step 4 is compared with actual ground friction factor, if required described ground friction coefficient is less than described actual ground friction factor, then described bipod walking robot walking is non-slip, otherwise then skids.

4. the walking control method of bipod walking robot as claimed in claim 1, it is characterized in that: if described bipod walking robot walking is non-slip in step 4, then the described Walking Mode of planned described bipod walking robot meets ZMP stability requirement.

5. the walking control method of bipod walking robot as claimed in claim 1, is characterized in that: the walking model that the generation of the described Walking Mode of bipod walking robot described in step 3 is is bipod walking robot with desk-dolly.

6. the walking control method of bipod walking robot as claimed in claim 5, is characterized in that: the generation of the described Walking Mode of bipod walking robot described in step 3 is for system inputs with the differential of the acceleration of barycenter.

7. the walking control method of bipod walking robot as claimed in claim 6, it is characterized in that: the generation of the described Walking Mode of bipod walking robot described in step 3 is that the parameter in evaluation function is for setting described walking parameter by the minimized method tracing control of evaluation function with reference to ZMP track.

8. the walking control method of bipod walking robot as claimed in claim 7, is characterized in that: the parameter in described evaluation function is for setting walking speed and the step-length of described walking parameter.

9. the walking control method of bipod walking robot as claimed in claim 8, it is characterized in that: the described Walking Mode of the described bipod walking robot generated in step 3 meets the ZMP stability criterion of double feet walking, combine with stability criterion of whether skidding.