CN104765372B - A kind of multi-foot robot keeps the gait planning method of linear translation - Google Patents

Abstract

The invention discloses a gait planning method for a multi-legged robot to maintain straight-line translation, which mainly solves the existing technical problems such as gait planning that easily leads to fuselage bumping. The main points of the technical solution are: two legs on one of the diagonal lines of the robot move forward by a distance S, the leg turns backward dγ once, the other legs are vertical, forward forward dγ once, and this process is repeated. Up to the forward motion unit distance (γ=γ _max ) the legs are transformed into the state perpendicular to the body (γ=0), and all the legs perpendicular to the body are transformed into the state of the backward movement unit distance (γ=-γ _max ), move all legs rotated to ‑γ angle forward two unit distances (that is, rotate to γ angle), complete a cycle of motion, and then perform a cycle cycle. It is mainly used for gait planning of multi-legged robots to maintain linear translation.

Description

A gait planning method for a multi-legged robot to maintain linear translation

technical field

The invention relates to a gait planning method for a multi-legged robot to maintain linear translation.

Background technique

At present, the walking gait of the robot determines the performance, stability, and working range of the robot. If the robot's gait is not planned properly, it will not be able to achieve the intended function. At present, there are many methods for gait planning of multi-legged robots, such as dog-like gait, turtle-like gait, single-step translational gait, etc. However, most of these gaits have some problems. For example, for a robot imitating a dog’s gait, the fuselage will rotate back and forth with the quadruped diagonal as the axis during walking, causing the fuselage to bump; for a robot imitating a turtle’s gait , the fuselage will rotate back and forth on the axis perpendicular to the upward direction of the fuselage; although the single-step translational gait will not rotate, the fuselage will bump up and down and so on. Based on the above examples, it can be seen that the various existing gait planning methods will generate other additional motions when the robot fuselage undergoes linear translation, and these additional motions will affect the performance of the robot. For example, if a camera is installed on the body of the robot to capture the front view, if the movement of the robot has other additional motions in addition to linear translation, it will cause the camera to shake and cause problems such as blurred images. To sum up, the existing various gaits are difficult to solve the impact of unnecessary additional motion on the motion of multi-legged robots during the robot motion process.

Contents of the invention

The purpose of the present invention is to provide a gait planning method for the multi-legged robot to maintain linear translational movement without other additional motions during the forward process of the multi-legged robot.

The technical solution adopted by the present invention to solve the technical problem is: the gait planning method is periodic, and the robot linear translation unit distance in each motion cycle, the specific steps are as follows:

Step 1: first give h, l ₀ constants; where: l ₀ is the distance between the landing point of each leg of the robot and the center of the fuselage on the horizontal plane, and h is the vertical height between the landing point and the fuselage;

Step 2: Move the two legs on one of the diagonal lines of the robot forward for a unit distance S, where γ is the acute angle between the projection l and l ₀ , and l is the position of a, b, and c on the same foot. The total length of the projection on the horizontal plane, a is the length of the cross arm 2, b is the length of the oblique arm 3, c is the length of the vertical arm 4, and γ _max is the maximum value of γ;

S＝l ₀ tanγmax——————————(1)

When γ=γ _max , the other two legs are perpendicular to the fuselage, that is, γ=0, while ensuring that the vertical height between the landing point of each leg and the fuselage is constant h and the landing point of each leg of the robot on the horizontal plane and the fuselage The distance between the centers is a constant l ₀ , that is, each moment has the following relationship:

bsinα+csin(α+β)=h—————(2)

bcosα+ccos(α+β)=l _c —————(3)

Among them, l _c is the projection of the upper oblique arm and the vertical arm on the horizontal plane, that is, l _c = la; β is the angle between the extension line of the oblique arm and the vertical arm, and α is the angle between the extension line of the oblique arm and the oblique arm horn;

Step 3: The robot starts to move in a straight line. In step 2, the leg that steps forward at an angle of γ rotates backward dγ for a single time. At the same time, all legs perpendicular to the fuselage in step 2 rotate forward dγ for a single time. This process is repeated. Until the forward movement unit distance S in step 2, that is, the leg with γ=γ _max is transformed into a state perpendicular to the fuselage, that is, γ=0, and all legs perpendicular to the fuselage in step 2 are transformed into a backward movement unit distance The state of , that is, γ=-γ _max , in the process, the angles of the other two joints are obtained with the change of the independent variable γ _i :

Perform error judgment on the instantaneous values γ _i , β _i , and α _i at each moment, remove the data groups that exceed the allowable range of the maximum error, and input the data groups that meet the requirements to the corresponding joints;

Step 4: Move all the legs rotated to -γ angle in step 3 forward by two unit distance S, that is, rotate to γ angle;

Step 5: Complete the single cycle and enter the next cycle.

At any moment in the above robot gait planning process, the following relationship exists:

γ _rf = γ _lb ; γ _lf = γ _rb = γ _max -γ _lf ———————(6)

Linear translational trajectory of the robot:

Among them, γ _lf : the γ angle corresponding to the left front foot of the robot; γ _rf : the γ angle corresponding to the right front foot of the robot;

γ _lb : the γ angle corresponding to the left rear foot of the robot; γ _rb : the γ angle corresponding to the right rear foot of the robot;

To sum up, γ _rf , γ _lb , γ _lf , γ _rb are obtained from formula (6)(7) in the process of single linear translation, and then the α, β, and the data group beyond the error range is eliminated by error analysis, so as to obtain the rotation angles of all joints required to maintain the linear translation of the fuselage at this moment.

According to the criterion of the linear translational motion of the multi-legged robot, it is judged whether the obtained gait trajectory data is within the actual error tolerance, and the data exceeding the error range is eliminated. The forward direction of the vertical robot is the X-axis direction to the right, and the allowable errors in the Z-axis and X-axis directions are ε _z and ε _x respectively;

The basis for judging the error is: at any moment when the robot moves, the displacement offset of the fuselage in other directions other than the linear translation direction is zero, thus obtaining the error under the current data of the Z-axis and the X-axis:

z _zmp = bsiα+csin(β+α)-n—————(8)

x _zmp ＝(bcosα+ccos(β+α)+a)cosγ-l ₀ ——(9)

If |Z _zmp |≤ε _z and |X _zmp |≤ε _x , the obtained data is within the allowable range of error, and input it into the robot to complete the corresponding gait adjustment; if |Z _zmp |≤ε _z , |X If any one of _zmp |≤ε _x is not established, the obtained data exceeds the error range, and it will be eliminated to ensure a stable gait.

The invention has the beneficial effects of ensuring only linear translational motion of the fuselage, eliminating incidental motion in other directions to the greatest extent, and avoiding bumping of the fuselage.

Description of drawings

Figure 1 is a schematic diagram of the structure of a quadruped robot with three degrees of freedom per leg.

Fig. 2 is a front view of Fig. 1 .

FIG. 3 is a top view of FIG. 1 .

FIG. 4 is a side view of FIG. 1 .

Fig. 5 is a schematic diagram of main parameters in robot motion planning.

detailed description

The present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments.

Since the present invention involves a robot with 2n (n=2, 3, 4 . . . ) legs, the following implementations will be described by taking a common quadruped robot with three degrees of freedom per leg as an example. Other robots that meet the requirements can be implemented according to this plan.

This multi-legged robot includes a fuselage, four mechanical legs, and each leg includes a cross arm 2, an oblique arm 3, and a vertical arm 4. In Fig. 1, the fuselage 1 is connected with the cross arm 2 of each leg, and the cross arm 2 is placed horizontally and can rotate around the axis in the vertical upward direction; the other end of the cross arm 2 of each leg is connected with one end of the oblique arm 3 , can rotate around the vertical direction of the plane formed by the two arms; the other end of the oblique arm 3 of each leg is connected with the upper end of the vertical arm 4, and can rotate around the vertical direction of the plane formed by the two arms; the vertical arm 4 The lower end of the joint touches the ground, and the cross arm 2, the oblique arm 3 and the vertical arm 4 are respectively driven by their corresponding joint batteries.

The gait planning is realized based on the above robot model.

As shown in Figure 1 and Figure 5, the variables involved are:

a: horizontal arm; b: oblique arm; c: vertical arm;

l: total projection length of a, b, c on the same foot on the horizontal plane;

l ₀ : the shortest total length of the projection l (at this time l ₀ is perpendicular to the fuselage);

S: the distance of a single translation of the robot;

l _c : the projection of the upper oblique arm b and the vertical arm c on the horizontal plane, that is, l _c = la;

Δl: the difference between the longest projection of projection l and l ₀ ;

h: The visible height between the leg landing point and the fuselage, that is, the total height of the cross arm a, oblique arm b, and vertical arm c in the vertical direction (the vertical height of the cross arm is zero);

β: Angle between the extension line of the oblique arm and the vertical arm;

α: Angle between the extension line of the cross arm and the oblique arm;

γ: the acute angle between the projection and the shortest projection l ₀ of L;

γ _lf : the γ angle corresponding to the left front foot of the robot;

γ _rf : the γ angle corresponding to the right front foot of the robot;

γ _lb : the γ angle corresponding to the left rear foot of the robot;

γ _rb : the γ angle corresponding to the right rear foot of the robot;

γ _max : γ maximum value.

1. Operation process

A multi-legged robot gait planning method that maintains linear translation The gait planning method described has periodicity, and the robot linear translation unit distance in every motion cycle, the specific steps are as follows:

(1). The vertical distance l ₀ between the landing point of each leg of the robot and the fuselage, and the vertical height h between the landing point and the fuselage are constant. Therefore, h, l ₀ constants are given first.

(2). Move the two legs on one of the diagonal lines of the robot forward for a unit distance

S＝l ₀ tanγ _max ——————————(1)

When γ=γ _max , the other two legs are perpendicular to the fuselage, that is, γ=0, while ensuring that the vertical height between the landing point of each leg and the fuselage is constant h and the landing point of each leg of the robot on the horizontal plane and the fuselage The distance between the centers is constant l ₀ . That is, each moment has the following relationship:

bsinα+csin(α+β)=h—————(2)

bcosα+ccos(α+β)=l _c —————(3)

(3). The robot starts to move in a straight line, and the leg that steps forward at an angle of γ in (2) turns backward dγ once, and at the same time, all legs perpendicular to the fuselage in (2) turn forward forward dγ once. This process is repeated until the unit distance of forward movement in (2), that is, the leg of γ=γ _max is transformed into a state perpendicular to the fuselage, that is, γ=0, and all legs perpendicular to the fuselage in (2) are transformed into The state of backward movement unit distance is γ=-γ _max . In this process, the angles of the other two joints are obtained with the change of the independent variable γ _i :

Perform error judgment on the instantaneous values γ _i , β _i , and α _i at each moment, remove the data groups that exceed the allowable range of the maximum error, and input the data groups that meet the requirements to the corresponding joints;

(4). Move all the legs rotated to the -γ angle in (3) forward by two unit distance S, that is, rotate to the γ angle;

(5). The single cycle is complete and enters the next cycle.

At any moment in the above robot gait planning process, the following relationship exists:

γ _rf = γ _lb ; γ _lf = γ _rb = γ _max -γ _lf ———————(6)

Linear translational trajectory of the robot:

To sum up, γ _rf , γ _lb , γ _lf , γ _rb are obtained from equations (6) (7) during a single linear translation. Then determine the α and β of each leg by formula (4) (5), and eliminate the data group beyond the error range by error analysis. Thus, at this moment, the rotation angles of all joints required to maintain the linear translation of the fuselage are obtained.

2. Error judgment

According to the criterion of the linear translational motion of the multi-legged robot, it is judged whether the obtained gait trajectory data is within the allowable actual error, and the data exceeding the error range are eliminated. (The direction of the Z-axis is defined as the upward direction of the vertical fuselage, and the direction of the X-axis is defined as the forward direction of the vertical robot on the horizontal plane. The allowable errors of the directions of the Z-axis and the X-axis are ε _z and ε _x respectively).

The error judgment basis is: at any moment when the robot moves, the displacement offset of the fuselage in other directions except the linear translation direction is zero. Thus, the error under the current data of the Z-axis and the X-axis is obtained:

z _zmp = bsinα+csin(β+α)-h—————(8)

x _zmp ＝(bcosα+ccos(β+α)+a)cosγ-l ₀ ——(9)

If |Z _zmp |≤ε _z and |X _zmp |≤ε _x , the obtained data is within the allowable range of error, and input it into the robot to complete the corresponding gait adjustment; if |Z _zmp |≤ε _z , |X If any one of _zmp |≤ε _x is not established, the obtained data exceeds the error range, and it will be eliminated to ensure a stable gait.

Claims (3)

1. a kind of multi-foot robot keeps the gait planning method of linear translation, it is characterized in that：Described gait planning method tool Have periodically, robot linear translation unit distance, is comprised the following steps that in each period of motion：

Step 1：First give h, l₀Constant；Wherein：l₀Between touchdown point and fuselage center for every leg of robot on horizontal plane Distance, vertical heights of the h between touchdown point and fuselage；

Step 2：Two legs of the robot wherein on a diagonal are moved forward into unit distance S, wherein, γ for projection l with l₀Between acute angle, l is the projection overall length of a, b, c in the horizontal plane on same pin, and a be transverse arm length, and b is oblique arm length Degree, c are vertical arm length, γ_maxFor γ maximum；

S=l₀tanγ_max—————————(1)

As γ=γ_max, other two legs are vertical with fuselage, i.e. γ=0, while ensure the every touchdown point of leg and hanging down for fuselage Straight height is that the distance between the touchdown point of every leg of robot and fuselage center are constant l on constant h and horizontal plane₀, i.e., it is every There is following relation at the individual moment：

Bsin α+csin (alpha+beta)=h --- --- (2)

Bcos α+ccos (alpha+beta)=l_c—————(3)

Wherein, l_cFor oblique arm b on a pin and the projections of vertical arm c in the horizontal plane, i.e. l_c=l-a；β is oblique arm extended line with erecting The angle of arm, α are the angle of transverse arm extended line and oblique arm；

Step 3：Robot starts linear translation, and the leg single for stepping γ angles in step 2 forward turns round d γ backward, simultaneously All leg singles vertical with fuselage rotate forward forward d γ in step 2, circulate this process, until the unit that travelled forward in step 2 Distance S, i.e. γ=γ_maxLeg be transformed into the state vertical with fuselage, i.e., the institute vertical with fuselage in γ=0, and step 2 There are the state that leg is transformed into motor unit distance backward, i.e. γ=- γ_max, in the process with independent variable γ_iChange obtain The angle in other two joints：

<mrow> <msub> <mi>&amp;beta;</mi> <mi>i</mi> </msub> <mo>=</mo> <mi>arccos</mi> <mrow> <mo>(</mo> <mfrac> <mrow> <msup> <mi>h</mi> <mn>2</mn> </msup> <mo>+</mo> <msup> <mrow> <mo>(</mo> <mi>l</mi> <mo>-</mo> <mi>a</mi> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mo>-</mo> <msup> <mi>c</mi> <mn>2</mn> </msup> <mo>-</mo> <msup> <mi>b</mi> <mn>2</mn> </msup> </mrow> <mrow> <mn>2</mn> <mi>b</mi> <mi>c</mi> </mrow> </mfrac> <mo>)</mo> </mrow> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>4</mn> <mo>)</mo> </mrow> </mrow>

<mrow> <msub> <mi>&amp;alpha;</mi> <mi>i</mi> </msub> <mo>=</mo> <mi>arctan</mi> <mrow> <mo>(</mo> <mfrac> <mi>h</mi> <mrow> <mi>l</mi> <mo>-</mo> <mi>a</mi> </mrow> </mfrac> <mo>)</mo> </mrow> <mo>-</mo> <mi>arccos</mi> <mrow> <mo>(</mo> <mfrac> <mrow> <msup> <mi>b</mi> <mn>2</mn> </msup> <mo>+</mo> <msup> <mrow> <mo>(</mo> <mi>l</mi> <mo>-</mo> <mi>a</mi> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mo>+</mo> <msup> <mi>h</mi> <mn>2</mn> </msup> <mo>-</mo> <msup> <mi>c</mi> <mn>2</mn> </msup> </mrow> <mrow> <mn>2</mn> <mi>b</mi> <msqrt> <mrow> <msup> <mrow> <mo>(</mo> <mi>l</mi> <mo>-</mo> <mi>a</mi> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mo>+</mo> <msup> <mi>h</mi> <mn>2</mn> </msup> </mrow> </msqrt> </mrow> </mfrac> <mo>)</mo> </mrow> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>5</mn> <mo>)</mo> </mrow> </mrow>

By the instantaneous value γ at each moment_i、β_i、α_iError judgment is carried out, the data group more than worst error allowed band is picked Remove, and satisfactory data group is inputed into corresponding joint；

Step 4：Rotation in step 3 to all legs at-γ angles is travelled forward two unit distance S, that is, rotated to γ angles；

Step 5：Single cycle is complete, into next cycle.

2. multi-foot robot according to claim 1 keeps the gait planning method of linear translation, it is characterized in that：In the above Any time in robot gait planning process is with the presence of following relation：

γ_rf=γ_lb；γ_lf=γ_rb=γ_max-γ_lf———————(6)

Robot linear translation track：

<mrow> <msubsup> <mo>&amp;Integral;</mo> <mn>0</mn> <msub> <mi>&amp;gamma;</mi> <mi>max</mi> </msub> </msubsup> <mi>l</mi> <mi> </mi> <mi>cos</mi> <mi>&amp;gamma;</mi> <mi>d</mi> <mi>&amp;gamma;</mi> <mo>=</mo> <msubsup> <mo>&amp;Integral;</mo> <mn>0</mn> <mi>S</mi> </msubsup> <mi>d</mi> <mi>S</mi> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>7</mn> <mo>)</mo> </mrow> </mrow>

Wherein, γ_lf：γ angles corresponding to robot left front pin；γ_rf：γ angles corresponding to robot right front pin；

γ_lb：γ angles corresponding to robot left back pin；γ_rb：γ angles corresponding to robot right back pin；

In summary, γ is obtained by formula (6) (7) during single linear translation_rf, γ_lb, γ_lf, γ_rb, then by formula (4) (5) α, β of every leg are determined, and the data group beyond error range is rejected by error analysis, thus obtains current time holding fuselage The articulate rotational angle of institute required for linear translation.

3. multi-foot robot according to claim 1 or 2 keeps the gait planning method of linear translation, it is characterized in that：Root Obtained gait track data is judged whether within actual error permission according to the criterion of multi-foot robot linear translation motion, And by the data rejecting beyond error range, wherein, with vertical fuselage upwards for Z-direction, with crossmachine people on horizontal plane Direction of advance is X-direction to the right, and Z axis and X-direction allowable error are respectively ε_z、ε_x；

The foundation of the error judgment is：At the time of any robot motion, other sides of the fuselage in addition to linear translation direction To shift offset be zero, the error being derived under Z axis and X-axis current data：

z_zmp=bsin α+csin (β+α)-h --- --- (8)

x_zmp=(bcos α+ccos (β+α)+a) cos γ-l₀——(9)

If | Z_zmp|≤ε_zAnd | X_zmp|≤ε_xThen the data obtained is in error allowed band, and is entered into robot, complete Into corresponding gait adjustment；If | Z_zmp|≤ε_z、|X_zmp|≤ε_xThere is any one invalid data that then obtain to exceed error range, And rejected to ensure that gait is steady.