JP3213897B2 - Automatic deviation correction method and apparatus for excavator and excavator speed control method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION The present invention relates to an excavator used for excavation of an excavation wall trench for constructing an underground continuous wall or the like, and when the excavator is vertically lowered, the deviation of the excavator from a target (partial deviation) is considered. The present invention relates to an excavator automatic deviation correcting method and apparatus capable of automatically correcting the excavation position, and an excavating speed control method capable of appropriately controlling the excavating speed of the excavator.

[0002]

2. Description of the Related Art For example, when constructing an underground continuous wall used for an earth retaining wall, a foundation of a structure, or the like, an excavation wall groove is formed in the ground on which the underground continuous wall is constructed, and the excavation wall groove is formed in the excavation wall groove. This is done by inserting a steel cage and placing concrete. In the case of excavating the excavation wall groove, usually, a lower end of the excavator body is provided by suspending and descending a horizontal multi-axial excavator provided with a drum-shaped excavator rotating around a horizontal axis. ing. In order to ensure the vertical accuracy of the excavation wall groove, it is necessary to excavate the excavator by vertically descending the excavator precisely.However, the excavator depends on the condition of the ground, the excavation conditions of the excavator itself, and the like. The deviation is corrected during excavation to deviate from the target.

Conventionally, the deviation correction is performed by providing images, graphs, character information, and the like by an underground continuous wall excavation accuracy management system (system) to an operator in real time.
This is performed by selecting a plurality of deflection correcting plates. The plurality of displacement correcting plates are attached to hydraulic jacks attached to a plurality of locations of the excavator, and these hydraulic jacks are appropriately expanded and contracted in the horizontal direction, so that the displacement correcting plates contact the wall surface of the excavation wall groove. By contact, the excavator is horizontally and rotationally moved in the excavation wall groove to correct the excursion of the excavator.

The above selection operation is performed as follows. (1) The operation pattern of the deviation correcting plate is determined from the deviation amount and the rate of change of the deviation amount. By repeating these selection operations several times, an appropriate operation pattern and operation amount for a certain period are determined. (2) Despite the above operation, if the effectiveness of the deflection correcting plate changes, a new operation pattern
Deal with the amount of operation. (3) When the deviation does not decrease despite the deviation correction operation, or from other observable factors (measurement data, etc.), the lowering speed of the excavator (or penetrating input) Reduce operation. (4) An operation of increasing the lowering speed (or penetrating force) of the excavator is performed based on a stable and small displacement amount and other observable factors (measurement data and the like).

[0005]

However, in the selection operation performed by the operator as described above, the eccentricity correction plate is not suitable because of the difference in the skill and skill of the operator and the ambiguity of the judgment criteria by the operator. Therefore, it was difficult to ensure uniform quality (excavation accuracy, excavation work period, etc.), and a long training period was required. In addition, when it is not possible to obtain an appropriate operation pattern and operation amount of the deviation correcting plate due to a lack of skill of the operator,
It was difficult to control the excavation speed of the excavator properly.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and provides an excavator capable of automatically correcting a deviation (deviation) from a target of an excavator without being affected by the skill of an operator. An object of the present invention is to provide an automatic deviation correction method and apparatus, and an excavation speed control method capable of appropriately controlling the excavation speed of an excavator.

[0007]

In order to achieve the above object, an automatic excursion correcting method for an excavator according to the present invention comprises:
Calculating the amount of deviation of the position of the excavator with respect to the control target during excavation and the rate of change of the amount of deviation, and performing fuzzy inference on the amount of deviation and the rate of change of the amount of deviation; A displacement correction control process that is provided in the excavator and controls the expansion and contraction of the jack portion that brings the excavator closer to the target by extending and contracting, every predetermined short period, and an average value of the deviation amount and the deviation amount And a gain adjustment control step of performing a fuzzy inference on the positive / negative sign reversal frequency of the above and controlling the expansion and contraction of the jack portion for each long cycle longer than the short cycle.

According to a second aspect of the present invention, there is provided an automatic excursion correcting device for an excavator, comprising: a measurement operation processing unit for calculating an excursion amount of a position of an excavator with respect to a control target during excavation and a rate of change of the excursion amount. By performing fuzzy inference on the amount of deviation and the rate of change of the amount of deviation, the excavator is provided in the excavator, and expands and contracts the jack portion that brings the excavator closer to the target. A deviation correction control unit that controls the average value of the deviation amount, and fuzzy inference on the positive and negative sign reversal degrees of the deviation amount, so that the expansion and contraction of the jack portion is increased at each long period longer than the short period. And a gain adjustment control unit for controlling.

[0009] The excavating speed control method for an excavator according to claim 3 comprises:
When excavating by the excavator while correcting the excursion of the excavator by the excavator automatic deviation correcting method according to claim 1, an upper limit value and a lower limit value for the body inclination angle and the amount of deviation of the excavator. If the excavator is deviated within a control value between the upper limit value and the lower limit value, the excavating speed of the excavator is increased, and the excavation speed is set outside the control value. When the excavator is displaced, the excavating speed of the excavator is reduced.

[0010]

According to the first and second aspects of the present invention, the amount of deviation of the excavator position with respect to the control target during excavation and the rate of change of the deviation amount are subjected to fuzzy inference to expand and contract the jack portion. Is controlled every predetermined short period,
By performing fuzzy inference on the average value of the deviation amount and the sign reversal frequency of the deviation amount, and controlling the expansion and contraction of the jack unit for each long period longer than the short period, the deviation correction control unit Evaluate the control performance that changes from time to time from the control result, adjust the gain by the gain adjustment control unit,
The operation amount (the amount of expansion / contraction operation of the jack unit) derived from the deviation correction control unit is corrected.

According to the third aspect of the present invention, an upper limit value and a lower limit value are set for the body inclination angle and the amount of deviation of the excavator, and a control value between the upper limit value and the lower limit value is set. If the excavator is deviated to, increase the excavation speed of the excavator,
Further, when the excavator is displaced outside the control value, the excavating speed of the excavator is reduced to increase the excavation speed when the transition of the amount of deviation is extremely stable. If the deviation is outside the control value (outside the upper and lower limits) while shortening the excavation time, the excavation speed is reduced and the correction of the deviation is quickly restored.

[0012]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing an embodiment of an excavator automatic displacement correcting method and apparatus according to the present invention;
First, the entire excavator including the excavator will be described with reference to FIGS. In FIG. 1, reference numerals 2A and 2B denote slide bases, which are arranged side by side in the X-axis direction (in the horizontal direction in FIG. 1) on both sides of the excavation site. On the upper surfaces of the slide bases 2A and 2B, support tables 3A and 3B
Are fixed respectively. The supports 3A and 3B have
The slide tables 4YA and 4YB are movably supported in the Y-axis direction (the direction orthogonal to the plane of FIG. 1), and the slide tables 4YA and 4YB have a slide table 4X.
A and 4XB are movably supported in the X-axis direction.

The slide bases 2A and 2B are provided with wire reels 10 on which a wire 8 for suspending the excavator body 6 of the excavator 5 is wound.
The leading ends of the wires 8 wound on these wire reels 10 pass through the fixed sheaves 12 on the lower surfaces of the slide tables 4XA and 4XB and the return sheaves 14 on the upper part of the excavator body 6 and slide tables 4XA and 4XB.
Respectively. Upper inclinometers 16A and 16B are provided at fixed positions of the wire 8 and the slide tables 4XA and 4XB, respectively, so that the inclination (angle with respect to the vertical direction) of the wire 8 is measured.

The excavator body 6 excavates by rotating a drum cutter 17 provided at the tip.
On the four side surfaces of the excavator body 6, fixed guide plates 22 are provided at upper and middle portions, respectively. Each of the fixed guide plates 22 can be protruded from the side surface of the excavator body 6 by a predetermined distance by a drive mechanism (not shown). Thereby, the excavator body 6 is supported in the excavation groove. The posture of the drum cutter 17 is determined by adjusting the upper eccentricity correcting plate 18, the middle eccentricity correcting plate 19 attached to the fixed guide plates 22, and the lower eccentricity correcting attached to the lower side surface of the excavator body 6. The adjustment is made by adjusting the distance between the plate 20 and the excavation wall groove (gut) 21 by moving the plate 20 in and out in the horizontal direction. On the side of the excavator body 6, a plurality of fixed guide plates 22 projecting from the side of the excavator body 6 over a certain distance are provided. Reference numeral 24 denotes a main body inclinometer for measuring the posture of the excavator main body 6.

As shown in FIG. 2, each of the deflection correcting plates 18, 19, 20 is attached to a plurality of hydraulic jacks 25 attached to the excavator body 6, and these hydraulic jacks 25 are extended and contracted. By doing so, they can be independently moved in and out in the horizontal direction. In FIG. 2, upper and lower deflection correcting plates 18 and 20 are shown, and a middle deflection correcting plate 19 is omitted. Each of the hydraulic jacks 25 is connected to an oil tank 27 via a hydraulic pipe 26, and a hydraulic pump 28 and an electromagnetic valve 29 are attached to the hydraulic pipe 26. The terminals e, f, g, and h of the solenoid valve 29 are connected to the terminals e,
f, g, and h, and the sequencer 30 is connected to the input / output device 31. The input / output device 31 is connected to input type measuring instruments such as the upper inclinometers 16A and 16B, the main body inclinometer 24, and the depth meter 32, and is connected to an output type processing unit 33. In FIG. 2, reference numeral 32a denotes a winch control panel, reference numeral 32b denotes a speed commander, and respective terminals b, c, and d are connected to corresponding terminals of the sequencer 30, and the depth gauge 32 is It is connected to the input / output device 31.

The processing unit 33 calculates a deviation between the position of the excavator during excavation and a control target of the excavator and a rate of change of the deviation, and a measurement calculation processing unit 34;
A fuzzy control unit 35 and a central processing unit (CP) connected to the measurement operation processing unit 34 and the fuzzy control unit 35
U) 36. As shown in FIG. 3, the fuzzy control unit 35 performs fuzzy inference on the deviation amount and the change rate of the deviation amount, and is provided in the excavator 5. A displacement correction control unit 37 for controlling the expansion and contraction of the hydraulic jack (jack portion) 25 approaching the target every predetermined short period, and a fuzzy adjustment of the average value of the deviation amount and the sign reversal frequency of the deviation amount. And a gain adjustment control unit 38 that controls the expansion and contraction of the hydraulic jack 25 for each long cycle longer than the short cycle, and evaluates the control performance that changes every moment from the control result by the deviation correction control unit 37. Then, the gain is adjusted by the gain adjustment control unit 38 and the deviation correction control unit 37 is adjusted.
The operation amount derived from is corrected.

In the above-described excavator, the upper inclinometers 16A and 16B, the depth meter 32, and the main body inclinometer 24
The positions of the deviation control points A, B, C, D of the excavator 5 are calculated based on the angle data and the excavator length detected by
The calculated value is compared with the target value, and fuzzy inference as described later is performed to obtain the deviation correction plates 18, 19, 2
By inserting and removing 0, the excursion of the excavator 5 is corrected. FIG. 4 shows the positional relationship of the deflection correcting plates installed in the excavator 5, the positions of the deviation control points A, B, C, and D of the excavator 5, and the signs of the deviation directions of those points. It is the figure which showed the definition, and (1)-(12) has shown the deflection correction plate.

FIG. 5 shows a procedure for calculating the positions of the deviation control points A, B, C and D based on the angle data detected by the upper inclinometers 16A and 16B and the main body inclinometer 24. It will be described with reference to FIG. That is, the data obtained from the upper inclinometers 16A and 16B is measured by the data of the inclination angles of the wire in both the X and Y directions (the upper inclination angle θLX θLY on the left and the upper inclination angle θRX θRY on the right) and the main body inclinometer 24. Data of the inclination angle (ΘX ′, ΘY ′) of the main body of the drum cutter 17 and the coordinates of the point C at one end of the center axis of the drum cutter 17 (X (L) ), Y (L), and point D at the other end
Are calculated based on the following equation.

The positions of the points C and D are determined by the two points A and B at the upper end of the excavator body 16 calculated based on the inclination angle of the wire (specifically, the deviation amounts XL and L of the left and right return sheaves 14). YL and XR, YL) and the deviations x ', y' of the points C, D at both ends of the drum cutter 17 which are geometrically determined only by the inclination of the excavator body 6.

The deviation amounts of the points A and B are as follows: A (left side): X direction deviation amount XL = De × tan (θLX)
/ B Y direction deviation amount YL = De * tan (θLY) / b B (right side): X direction deviation amount XR = De * tan (θRX)
/ D Y-direction deviation amount YR = De × tan (θRY) / d.

The deviation amounts of the points C and D at both ends of the drum cutter 17 caused only by the inclination of the excavator body 6 are as follows:
In the X ′ and Y ′ coordinates rotated by −α degrees with respect to the X and Y coordinates, X′-direction displacement x ′ = L × tan (ΘX ′) / f Y′-direction displacement y ′ = L × tan (ΘY ′) / f The amount of displacement in the combined direction z ′ = L × e / f.

Where θLX and θLY are the left upper tilt angles θRX and θRY are the right upper tilt angles ΘX 'and ΘY' are the main body tilt angles L is the total length of the excavator main body De is the displacement calculation depth, and , A = {tan ² (θLX) + tan ² (θLY)} ^1/2 b = {tan ² (θLX) + tan ² (θLY) +1} ^1/2 c = {tan ² (θRX) + tan ² (θRY)} ^1/2 d = ｛tan ² (θRX) + tan ² (θRY) +1｝ ^1/2 e = ｛tan ² (ΘX ') + tan ² (ΘY')｝ ^1/2 f = ｛tan ² (ΘX ') + tan ² (ΘY ') + 1｝ ^1/2 .

Here, a point (x ′, Y ′) on the X ′, Y ′ coordinates
y ′) is converted into X, Y coordinates obtained by rotating clockwise (clockwise) α degrees, and (x, y), and the length of each line segment in FIG. OR = AB =
Assuming that CD = P (P = 2900 mm in the embodiment), cos α = (P−XL + XR) / P or sin α =
Since (YL−YR) / P, α = cos ⁻¹ {(P−XL + XR) / P} = sin
^-1 {(YL-YR) / P} x = x ′ × cos α + y ′ × sin α y = −x ′ × sin α + y ′ × cos α Therefore, the coordinates of both ends C and D of the rotation center axis of the drum cutter 16 in the XY coordinate system having the origins at the left and right points OL and OR, that is, the deviation amount are as follows. X (L) = XL + x = De × tan (θLX) / b + x ′ ×
cosα + y ′ × sinα Y (L) = YL + y = De × tan (θLY) / b−x ′ ×
sinα + y ′ × cosα X (R) = XR + x = De × tan (θRX) / d + x ′ ×
cosα + y ′ × sinα Y (R) = YR + y = De × tan (θRY) / d−x ′ ×
sinα + y '× cosα

As described above, each control point A,
The deviation amounts of B, C, and D can be known. Next, for each of the control points A, B, C, and D, control is performed by performing fuzzy inference. Before that, the physical quantities (control quantities and operation quantities) are standardized as follows. Deviation of deviation The deviation is equal to the deviation since the target value of the controlled variable is zero (R0 = 0). In the initial state, the deviation amount is made to correspond to the following standard values. Deflection amount: -100mm 0mm + 100mm Variable setting Standard value: -1.0 0 +1.0 Rate of change of deviation of deviation The change rate of deviation in the initial state corresponds to the following standard value. Change rate of deviation amount: -50mm 0mm + 50mm Variable setting Standard value: -1.0 0 +1.0 Operation amount The operation amount corresponds to the following standard value in the initial state. Operation amount: -15sec 0sec + 15sec Setting variable Standard value: -1.0 0 +1.0 However, positive operation amounts are: (3), (4), (7), (8), (9),
(11) Negative manipulated variables are as follows: Supply hydraulic fluid to hydraulic cylinders for (1), (2), (5), (6), (10), and (12), and displace each oil. It is assumed that the amount correction plate is extended.

During the excavation, the excursion amount is obtained at regular intervals, and the rate of change of the excursion amount is obtained.
Apply fuzzy inference to them. The fuzzy inference will be described below. The figures shown in FIGS. 6 to 8 are respectively a displacement amount, for example, a membership function of (YL), a membership function of the rate of change of displacement (ΔYL), and a membership function of increment of operation amount (ΔuLY). An example is shown below. In FIG. 9, NB, NM,
NS, Z0, PS, PM, and PM are the deviation amounts (Y
L), a fuzzy set of the rate of change of the amount of deviation (ΔYL). Each fuzzy set includes NB: negatively large, NM: negatively medium, NS: negatively small, Z0: almost zero, P0
Values corresponding to S: positively small, PM: positively medium, and PB: positively large belong to each.

Now, consider a case where the standard value of the deviation amount (YL) is α and the standard value of the rate of change of the deviation amount (ΔYL) is β. In this case, the displacement (YL) can be an element of either the fuzzy set NM or NB, and the rate of change of the displacement (ΔYL) can be an element of either the fuzzy set NS or NM. Therefore, as a case to be calculated, (a) when the displacement amount (YL) = NM and the rate of change of the displacement amount (ΔYL) = NS, (b) the displacement amount (Y
When L) = NM and the rate of change of the displacement (ΔYL) = NM, (c) When the quantity of the displacement (YL) = NB and the rate of change of the displacement (ΔYL) = NS, d) Deflection amount (YL) = N
B, and when the rate of change of the amount of deviation (ΔYL) = NM.

Then, the grade (fitness) of each of the above four cases is determined as follows. As shown in FIG. 10, in the linear equation of the membership function, when the horizontal axis is divided by points a to g, the linear equation of each section is as follows. In the section: a to b, the NB straight line is (a, 1), (b,
0) because it is a straight line passing through the two points, NB straight line, Y = {(0-1) / (ba)} * (Xa) +
1. Since the NM straight line is a straight line passing through the two points (a, 0) and (b, 1), the NM straight line, Y = {(1-0) / (ba)} * (Xa) +
0, section: b to c, NM straight line, Y = {(0-1) / (c−b)} * (X−b) +
1, NS straight line, Y = {(1-0) / (c−b)} * (X−b) +
0, section: NS straight line, from c to d, Y = {(0-1) / (dc)} * (Xc) +
1, Z0 straight line, Y = {(1-0) / (dc)} * (Xc) +
0, Section: From d to e, Z0 straight line, Y = {(0-1) / (ed)} * (Xd) +
1, PS straight line, Y = {(1-0) / (ed)} * (Xd) +
0, section: PS straight line, e = f, Y = {(0-1) / (fe)} * (Xe) +
1, PM straight line, Y = {(1-0) / (fe)} * (Xe) +
0, section: f-g, PM line, Y = {(0-1) / (g-f)} * (X-f) +
1, PB straight line, Y = {(1-0) / (g-f)} * (X-f) +
0,

Now, assuming that a <α <b in YL and b <β <c in ΔYL, the grade for YL is Y on the NB straight line.
Assuming α (NB), Yα (NB) = {(0-1) / (ba)} * (α-a) +
1. Assuming that the grade on the NM straight line is Yα (NM), Yα (NM) = {(1-0) / (ba)} * (α-a) +
0, (see FIG. 11). Assuming that the grade on the NB line is Yβ (NM), the grade for ΔYL is Yβ (NM) = {(0−1) / (c−b)} * (β−b) +
1. If the grade on the NS straight line is Yβ (NS), Yβ (NS) = {(1-0) / (c−b)} * (β−b) +
0 (see FIG. 12).

Next, in the above four cases, the respective grades (degrees of conformity) are compared, and the smaller value is adopted, and the increase (ΔuLY) of the operation amount corresponding to the smaller grade is determined. The area and the X coordinate of the center of gravity in the membership function are obtained. As shown in FIG. 10 and FIGS. 13 to 16, the membership function of the consequent part (the membership function of ΔuLY)
Is a triangle such as NM, NS, Z0, PS, PM, the X coordinate of the center of gravity is known because it is the X coordinate of the vertex of the triangle. That is, NM: X = b, NS: X = c, Z0: X = d (generally = 0), PS: X = e, and PM: X = f.

The area, that is, the area of the trapezoidal portion surrounded by the straight line of the membership function, the X axis, and the straight line indicating the value of the smaller grade is obtained as follows. In FIG. 17, since the points p, r, t and Yγ are known, the area of the trapezoidal portion can be obtained by knowing the X coordinates of the points q and s. The straight line A is X = Ｘ (rp) /
(1-0)｝ * (Y-0) + p, where Y
By setting = Yγ, the X coordinate of q can be obtained. X = q = {(rp) / (1-0)} * (Yγ-0) + p The straight line B is X = {(tr) / (0-1)} * (Y-1) + r
If Y = Yγ in this equation, the X coordinate of s can be obtained. X = s = {(tr) / (0-1)} * (Yγ-1) + r Therefore, the area S of the trapezoidal portion is as follows: S = (L1 + L2) * Yγ / 2 = ｛| s -q | + | tp |｝ *
Yγ / 2.

As shown in FIG. 18, the membership function of the consequent part (the membership function of ΔuLY) is NB,
In the case of a linear shape such as PB, the area of a trapezoidal portion surrounded by the membership function straight line, the X axis, and the straight line indicating the value of the smaller grade is S = 、 | r−q | + | Rp
|｝ * Yγ / 2. Note that the X coordinate of the point q can be obtained by calculating the equation of the straight line A or B in the same manner as described above, and by setting Y = Yγ in the equation. X = q = {(r−p) / (1−0)} * (Yγ−0) + p The triangles and the rectangular areas S1 and S2 in the trapezoid are S2 = | rq | *, respectively. Yγ, S1 = | q−p |, and letting the centroid points on the X axis of the rectangle and the triangle be n and m, respectively, L1 = (q−p) * 2/3, L2 =
{R- (rq) / 2} + (q-p). Therefore,
Assuming that the distance to the center of gravity G of the trapezoid is L3 (including the sign), S * L3 = S2 * L2 + S1 * L1 L3 = (S2 * L2 + S1 * L1) / S. Above,
This means that the conditions for calculating the conclusion part of the fuzzy inference have been met.

Next, a specific example will be described. As the antecedent condition, membership functions NB, NM, NS, Z0, P
The X coordinate (standard value) of each vertex of S, PM, and PB is determined as follows. a = -1.0, b = -2 / 3, c = -1 / 3, d = 0, e = + 1/3,
f = + 2/3, g = + 1.0 Also, the standard values α and β of the control amount are α = −0.80, β = −0.35
Assume that

(A) When the displacement (YL) = NM and the rate of change of the displacement (ΔYL) = NS (see FIG. 13), α
-NM line, grade Yα (NL
M) is: Yα (NM) = {(1-0) / (ba)} * (α-a) = +
0.6, in the β-NS line, grade Yβ against ΔYL
(NS) is given by: Yβ (NS) = {(1-0) / (c−b)} * (β−b) = +
0.95. Therefore, the smaller value Yα (NM) = + 0.
6 is adopted. On the other hand, the deviation amount (YL) is the fuzzy set N
If the parameter belongs to M and the rate of change of the deviation (ΔYL) belongs to the fuzzy set NS, PS is selected as the fuzzy set of the increment of the operation amount (ΔuLY) according to the table shown in FIG. Therefore, since p = + 2/3, r = + 1/3, and t = 0 in the PS function in the consequent part (the membership function of ΔuLY), q and s are: q = ｛(rp ) / (1-0)｝ * {Yα (NM) -0｝ + p =
+0.47 s = {(tr) / (0-1)} * {Yα (NM) -1} + r =
+0.2. Therefore, the area SA of the trapezoidal portion is: SA = ｛| sq | + | tp−｝ Yα (NM) * / 2 = (0.
27 + 0.76) * 0.6 / 2 = 0.28. The X coordinate of the center of gravity is X = e = + 1/3.

(B) When the displacement (YL) = NM and the rate of change of the displacement (ΔYL) = NM (see FIG. 14), α
-NM line, grade Yα (NL
M) is Yα (NM) = + 0.6, β-NM line, grade Yβ against ΔYL
(NM) is given by Yβ (NM) = {(0−1) / (c−b)} * (β−b) + 1 =
+0.05. Therefore, the smaller value Yβ (NM) = + 0.
05 is adopted. On the other hand, when the displacement amount (YL) belongs to the fuzzy set NM and the change rate of the displacement amount (ΔYL) belongs to the fuzzy set NM, the fuzzy set of the operation amount increment (ΔuLY) according to the table shown in FIG. Is selected as PM. Therefore, in the PM function in the consequent part, p = + 1.
Since 0, r = + 2/3 and t = + 1/3, q and s are given by: q = {(rp) / (1-0)} * {Yβ (NM) -0} + p =
+0.98 s = {(tr) / (0-1)} * {Yβ (NM) -1} + r =
+0.35. Therefore, the area SB of the trapezoidal portion
Is: SB = ｛| sq | + | tp−｝ Yβ (NM) * / 2 = (0.
63 + 0.67) * 0.05 / 2 = 0.033. Also,
The X coordinate of the center of gravity is X = f = + 2/3.

(C) When the displacement (YL) = NB and the rate of change of the displacement (ΔYL) = NS (see FIG. 15), α
-In the NB straight line, the grade Yα (NL
B) is given by Yα (NB) = {(0−1) / (ba)} * (α−a) + 1 =
In the +0.40 β-NS line, grade Yβ against ΔYL
(NS) is Yβ (NS) = + 0.95. Therefore, the smaller value Yα (NB) = + 0.0.
Adopt 40. On the other hand, when the displacement amount (YL) belongs to the fuzzy set NB and the rate of change of the displacement amount (ΔYL) belongs to the fuzzy set NS, the fuzzy set of the increment (ΔuLY) of the operation amount according to the table shown in FIG. Is selected as PM. Therefore, in the PM function in the consequent part, p = + 1.
Since 0, r = + 2/3 and t = 1/3, q and s are: q = {(rp) / (1-0)} * {Yα (NB) -0} + p =
+0.87 s = {(tr) / (0-1)} * {Yα (NB) -1} + r =
+0.47. Therefore, the area SC of the trapezoidal portion
SC = ｛| sq | + | tp−｝ Yα (NB) * / 2 = (0.
40 + 0.67) * 0.40 / 2 = 0.21. The X coordinate of the center of gravity is X = f = + 2/3.

(D) When the displacement (YL) = NB and the rate of change of the displacement (ΔYL) = NM (see FIG. 16), α
-In the NB straight line, the grade Yα (NL
B) is the grade Yβ against ΔYL in the Yα (NB) = + 0.40 β-NM straight line.
(NM) is given by Yβ (NM) = {(0−1) / (c−b)} * (β−b) + 1 =
0.05. Therefore, the smaller value Yβ (NM) = + 0.
05 is adopted. On the other hand, when the displacement amount (YL) belongs to the fuzzy set NB and the change rate of the displacement amount (ΔYL) belongs to the fuzzy set NM, the fuzzy set of the operation amount increment (ΔuLY) according to the table shown in FIG. Is selected as PB. Therefore, in the PB function in the consequent part, p = + 2 /
3, since r = + 1.0, q and s are given by q = ｛(rp) /
(1-0)｝ * ｛Yβ (NM) -0｝ + p = + 0.68. Therefore, the area SD of the trapezoidal portion is: SD = ｛| r−q | + | rp −｝ * Yβ (NM) / 2 = (0.
32 + 0.33) * 0.05 / 2 = 0.016. Now, the areas of the triangle and the rectangle in the trapezoid are S1 and S2, respectively.
Then, S1 = | q-p | * Yβ (NM) /2=0.0005, S2 = | rq | * Yβ (NM) = 0.016 p
Assuming that the distances from the point to the center of gravity of the triangle and rectangle on the X axis are L1 and L2, respectively, and the distance from the center of gravity of the trapezoid is L3, L3 = (S1 * L1 + S2 * L2) / S, L1 = ( q-p) *
Since 2/3 and L2 = {r- (rq) / 2} + (q-p), L1 = + 0.009, L2 = + 0.85, and L3 = + 0.85.

Thus, four sets of MIN conditions are completed. We will take MAX in the conclusion. In summary, in the case of (A), SA = 0.28, LA = + 1/3 (B), SB = 0.033, LB = + 2/3 (C), SC = 0.21 , LC = + 2/3 (D), SD = 0.016 and LD = + 0.85. Therefore, if the fuzzy variable (value of the center of gravity) of the output part of the membership function of the consequent part is X0, X0 = (SA * LA + SB * LB + SC * LC + SD * LD) /
S = + 0.5. Here, S = SA + SB + SC + SD. Finally, the inference with the standard value is returned to the physical quantity (operating quantity), and ΔuLY = + 15 sec * (+ 0.5) / (+ 1.0) = + 7.5
sec. This corresponds to supplying oil for 7.5 seconds in FIG. 4 to extend the deviation correcting plate (8) for the deviation control point A. Similarly, fuzzy inference is performed for the deviation amount control points B, C, and D, and control in both the X and Y directions is performed for all deviation amount control points. In the control at the displacement control points A, B, C, and D, at the positions of the control points A and C with respect to the target reference,
When the codes of A and C are the same, only the control at the control point C is performed, and the control at the control point A is ignored. Also,
When the codes of the control points A and C are different codes, control is performed at both control points A and C.

As described above, the displacement correction control makes use of the production rules and fuzzy inference to independently control the four displacement amounts in the Y-axis direction for the four displacement control points (A, B, C, D). Is determined, and 2 is set in the upper and lower X-axis directions.
One operation amount is determined. Learning control for gain adjustment is performed by evaluating the control performance that changes from time to time based on these control results, and the gain is adjusted for the four points described above. In the displacement correction control, the increment of the operation amount is determined every T1 seconds (in a cycle), and expansion / contraction control of the actuator (hydraulic cylinder for the displacement correction plate) is performed. The amount of gain increase / decrease is determined (periodically), and the manipulated variable derived from the deviation correction control is corrected. However, it is assumed that T2 >> T1.

The gain is 1 in the initial state. It is assumed that time has elapsed and the current time has become (k) and (i). At this time, the deviation of the final deviation correction operation amount including the learning control toward time (i + 1) is defined as Δ ″ uLY (i), and the deviation obtained in the T1sec cycle regardless of the learning control. ΔuLY represents the increment of the correction operation amount toward time (i + 1).
(I). The total gain derived at the previous learning control time (k-1) including the learning control is represented by g (k-1). This time, assuming that the gain increment obtained at time (k = i) toward time (i + 1) is Δg (k) and the total gain is g (k), Δ ″ uLY (i) = g (k ) * ΔuLY (i) = ｛g (k−
1) + Δg (k)｝ * ΔuLY (i)

Therefore, this time, at time (k = i), the total deviation correction operation amount including the learning control (gain adjustment control) toward time (i + 1) is set to uLY (i), and Assuming that the total deviation correction operation amount including the learning control toward time (i) at time −1) is uLY (i−1), uLY (i) = uLY (i−1) + Δ ″ uLY (I) = uL
Y (i) + {g (k-1) + Δg (k)} * ΔuLY (i) Therefore, g (k) = g (k−1) + Δg (k), and the learning control of gain adjustment is position-type control. At this time, the actual operation amount (output) of the deviation correction control is Δ ″ uLY (i) only in increments, and is a speed-type control. There are various evaluation methods for learning control. Then, the fuzzy inference is applied to the average value (correctly, the square root of the average value of the sum of squares: because there is a sign) and the sign inversion frequency of the deviation amount during T2 sec in the gain adjustment cycle period. Then:

The graphics shown in FIGS. 19 to 22 are
An example is shown in which the membership function of the average value (ψ) of the deviation amount, the membership function of the sign inversion frequency (M) of the deviation amount, and the membership function of the gain adjustment amount (Δg) are shown for each of the four cases. . In FIG. 23, NB, NS,
Z0, PB, and PS are fuzzy sets of gain adjustment amounts, each of which has the following meaning. NB: Gain is greatly reduced. NS: The gain is slightly reduced. Z0: Do not change the gain. PB: Increase the gain greatly. PS: Gain is slightly increased.

Here, the physical quantity is standardized. (A) Average value of deviation amount: Since the sign of the deviation amount is positive or negative, an average value of 10 times is defined as the following equation. p = ｛(ΣYL (i) ² ) / N｝ ^1/2 where N = 1
0 Now, if the following data is obtained, YL (1) = 15 mm, YL (2) =-10 mm, YL (3) = 32 mm,
YL (4) = 12mm, YL (5) =-28mm, YL (6) = 8mm, Y
L (7) =-6mm, YL (8) = 11mm, YL (9) =-16mm, Y
L (10) = 13 mm p = (2923/10) ^1/2 = + 17.1 mm Standard deviation value: +90 mm 0 mm Variable setting Standard value: +1.0 0 Therefore, in the above calculation example, the average The standard value a for the value is a = (+ 1.0) * (+ 17.1) / (+ 90) = + 0.1
9

(B) Degree of sign inversion of the amount of deviation: M The maximum number of sign inversions is nine, and the minimum number is zero. Sign inversion frequency: 9 times 0 times Setting change Standard value: +1.0 0 Therefore, if the sign inversion frequency is 4 times, the standard value a is a = (+ 1.0) * 4 / 9 = + 4.44
Becomes

(C) Gain adjustment amount: Δg Gain adjustment amount: + 0.5-0.5 Setting change Standard value: + 1.0-1.0 Therefore, when the gain adjustment amount is now +0.2 And the standard value a for it is a = (+ 1.0) * (+ 0.2) / (+ 0.0.
5) = + 0.4.

Now, as shown in FIGS. 19 to 22,
The following four cases are considered for the average value (ψ) of the deviation amount and the sign inversion frequency (M) of the deviation amount. (A) ψ = B and M = B (b) ψ = B and M = S (c) ψ = S and M = B (d) ψ = S and M = S

Now, the standard values of the control amounts are respectively the average value of the deviation amount: ψ (i) = + 0.82, the sign inversion frequency: M (i) =
Assume that it was +0.23. Then, the displacement amount (Y
L) and the rate of change in deviation (ΔYL) are determined in the same manner as when fuzzy inference is performed. In the case of (a), the grade Y for ψ is +0.
64, and the grade YM for M is +0.23. Therefore, the smaller value grade YM = + 0.2
3 is adopted. On the other hand, when ψ belongs to the fuzzy set B and M belongs to the fuzzy set B, NB is selected as the fuzzy set of the gain adjustment amount (Δg) according to the table shown in FIG. In addition, in the membership function of the gain adjustment amount (Δg) shown in FIG. 19, a trapezoidal area SA = 0.0989 surrounded by a straight line of the NB function, a straight line of grade YM = + 0.23, and the X axis. , The trapezoidal center of gravity LA = −0.738.

In the case of (b), grade Y for {
+0.64, and the grade YM for M is +0.77. Therefore, the smaller value grade Yψ = + 0.
64 is adopted. On the other hand, when ψ belongs to the fuzzy set B and M belongs to the fuzzy set S, PB is selected as the fuzzy set of the gain adjustment amount (Δg) according to the table shown in FIG. In the membership function of the gain adjustment amount (Δg) shown in FIG. 20, a trapezoidal area SB = 0.2176 surrounded by a straight line of the PB function, a straight line of grade Yψ = + 0.64, and the X axis is obtained. , And the trapezoidal center of gravity LB = −0.713.

In the case of (c), the grade Y for ψ is
+0.36, and the grade YM for M is +0.23. Therefore, the smaller value grade YM = + 0.
23 is adopted. On the other hand, when ψ belongs to the fuzzy set S and M belongs to the fuzzy set B, NS is selected as the fuzzy set of the gain adjustment amount (Δg) according to the table shown in FIG. In the membership function of the gain adjustment amount (Δg) shown in FIG. 21, the trapezoidal area SC = 0.2355 surrounded by the straight line of the NS function, the straight line of grade YM = + 0.23, and the X axis is obtained. And the trapezoidal center of gravity LC = −0.5.

In the case of (d), the grade Y for ψ is
+0.36, and the grade YM for M is +0.77. Therefore, the smaller value grade Yψ = + 0.
36 is adopted. On the other hand, when ψ belongs to the fuzzy set S and M belongs to the fuzzy set S, PS is selected as the fuzzy set of the gain adjustment amount (Δg) according to the table shown in FIG. Further, in the membership function of the gain adjustment amount (Δg) shown in FIG. 22, the trapezoidal area SD = 0.2952 surrounded by the straight line of the PS function, the straight line of grade Yψ = + 0.36, and the X axis. And the trapezoidal center of gravity LD = + 0.5.

Then, the MAX processing is performed at the conclusion part. Assuming that the fuzzy variable (value of the center of gravity) at the output of the membership function is X0, S * X0 = SA * LA + SB * LB + SC * LC + SD * LD
Therefore, (however, S = SA + SB + SC + SD) X0 = (SA * LA + SB * LB + SC * LC + SD * LD) /
S = + 0.157. Therefore, when the gain adjustment amount Δg is obtained, Δg = + 0.5 * (+ 0.157) / (+ 1.0) = + 0.0785
Becomes The gain of the previous time (before T2sec) is g (k-
1), and the current gain is g (k), g (k) = g (k-1) +0.0785, and the gain is slightly increased from the previous time.

Next, an example of a method for controlling the excavating speed of the excavator according to the present invention will be described. As described above, the deviation correction control and the gain adjustment control are performed. However, in spite of these controls, the deviation amount cannot be controlled within the management value and may deviate from the controllable state by the automatic driving. .
Eventually, the excavation speed is reduced (or the penetration force is reduced) while the automatic operation is being performed to recover the situation, or if the recovery is still unsatisfactory, the operation is switched to the manual operation and the excavation is advanced. If the transition of the deviation amount is extremely stable in the automatic operation state, the speed is increased (or the penetration force is increased) in order to shorten the excavation time.

In the excavation speed (penetration input) live value control, the upper and lower limits (both upper and lower limits are variable) are set (referred to as live values) for the following two items. A message "", "decelerate" or "increase the penetration force" or "decrease the penetration force" is issued to give an instruction to the operator. The operator receiving the instruction manually performs the acceleration / deceleration operation.

That is, as shown in FIG. 24, the upper limit of the control limit at the inclination angle (minute) of the excavator body is φx, φ
y, control limit lower limit values are ωx, ωy, control limit total deviation (mm): upper limit of control limit of deviation amount of deviation control points C and D is Φx, Φy, lower limit of control limit is Ωx, Ωy, each measured value is ΘX, ΘY, X (C), Y (C), X
Assuming that (D) and Y (D), (a) the condition for issuing the message “decelerate” or “decrease the input force” is “|「 x |> φx or | Θ
y |> φy ”or“ | X (C) |> Φx or | Y (C) |> Φy ”or“ | X (D) |> Φ
x or | Y (D) |> Φy ”. (B) The conditions for issuing the message “increase the speed” or “increase the penetration force” are as follows: || x | <ωx and | Θy
| <Ωy ”and“ | X (C) | <Ωx and | Y
(C) | <Ωy ”and“ | X (D) | <Ωx and | Y (D) | <Ωy ”. (C) When conditions other than the above conditions are satisfied, the status quo is maintained and no message is issued.

By controlling the excavating speed of the excavator in this manner, the excavating speed can be increased to shorten the excavating time when the change in the amount of deviation is extremely stable. When the deviation is outside the control value (outside the upper and lower limit values), the excavation speed is reduced, and the correction and recovery of the deviation can be quickly performed.

In the above embodiment, the fuzzy control is performed by using the triangular membership function. However, the change (tuning) of the membership function can be simplified as follows. That is, conventionally, in the fuzzy control package software on the market, the change of the triangular membership function has been performed by numerically inputting the coordinates of the end or the vertex of the base of the triangle to be changed. Tuning by simulation or actual machine required trial and error, and changing the membership function by numerical input took a great deal of effort and time. Also, due to the nature of fuzzy control, it is important to grasp and confirm the trend in the initial simulation or tuning, and in many cases, it is not necessary to set fine numerical values.

Therefore, when it is desired to change the shape of the triangle of the triangle type membership function, the cursor is placed on the end or the vertex of the base of the triangle, the point is determined (by pressing the return key), and the left and right arrows are set. The key (cursor key) is moved horizontally to the place you want to set, and after it is moved to that place, a new position is determined with the return key.

FIG. 25 shows an operation example. Now, in order to increase the sensitivity of the deviation control point (A) in the Y direction deviation amount, the function key f2 is selected. In order to make the base of the Z0 triangle smaller, the cursor can be moved to the left end of the base of the Z0 triangle with the arrow keys and the return key is pressed, so that fine adjustment can be performed by operating the arrow keys. If the arrow keys are continuously pressed, horizontal movement can be continuously performed. The state where the adjustment is finally completed is the lower diagram. By employing this method, the labor and time of the simulation operation can be greatly reduced.

[0058]

As described above, claims 1 and 2
According to the invention, during the excavation, the amount of deviation of the position of the excavator with respect to the control target, and the rate of change of the amount of deviation are subjected to fuzzy inference to control the expansion and contraction of the jack portion at predetermined short periods. Meanwhile, the average value of the deviation amount and the sign reversal frequency of the deviation amount are subjected to fuzzy inference, and the expansion and contraction of the jack portion is controlled for each long period longer than the short period. The control performance that changes every moment is evaluated from the control result by the control unit, and the gain is adjusted by the gain adjustment control unit, so that the operation amount (the expansion / contraction operation amount of the jack unit) derived from the deviation correction control unit can be corrected. . Therefore, the deviation (deviation) of the excavator from the target can be automatically and reliably corrected without being affected by the skill of the operator.

According to the third aspect of the present invention, when the excavator performs excavation while correcting the excursion of the excavator by the automatic excavator deviation correcting method, the body inclination angle of the excavator and An upper limit and a lower limit are set for the amount of deviation, and when the excavator is displaced within a control value between the upper limit and the lower limit, the excavating speed of the excavator is increased. Further, when the excavator is out of the control value, the excavating speed of the excavator is reduced, so that when the change in the amount of deviation is extremely stable, the excavating speed is reduced. The excavation time can be shortened by increasing the speed. On the other hand, when the deviation amount is outside the control value (outside the upper and lower limit values), the excavation speed is reduced,
Correction and recovery of the deviation can be performed promptly.

[Brief description of the drawings]

FIG. 1 is a longitudinal sectional view showing an embodiment of an excavator according to the present invention.

FIG. 2 is a configuration diagram showing one embodiment of a drilling device according to the present invention.

FIG. 3 is a block diagram of the control system.

FIG. 4 is a perspective view showing a position of a deflection correcting plate.

FIG. 5 is a view of the excavator suspended by wires as viewed from above.

FIG. 6 is a graph showing a deviation amount membership function.

FIG. 7 is a graph showing a membership function of a rate of change in the amount of deviation.

FIG. 8 is a graph showing a membership function of an operation amount increment.

FIG. 9 is a diagram showing a table of production rules.

FIG. 10 is a graph showing a membership function.

FIG. 11 is a graph showing a main part in FIG. 6;

FIG. 12 is a graph showing a main part in FIG. 7;

FIG. 13 is a graph showing a membership function.

FIG. 14 is a graph showing a membership function.

FIG. 15 is a graph showing a membership function.

FIG. 16 is a graph showing a membership function.

FIG. 17 is a graph showing the consequent part of the membership function.

FIG. 18 is a graph showing the consequent part of the membership function.

FIG. 19 is a graph showing a membership function.

FIG. 20 is a graph showing a membership function.

FIG. 21 is a graph showing a membership function.

FIG. 22 is a graph showing a membership function.

FIG. 23 is a diagram showing a table of production rules.

FIG. 24 is an explanatory diagram for explaining live value control.

FIG. 25 is an explanatory diagram for explaining a method of tuning a membership function.

[Explanation of symbols]

 5 Excavator 6 Excavator body 18, 19, 20 Deflection correction plate 25 Hydraulic jack (jack portion) 34 Measurement calculation processing part 37 Deflection correction control part 38 Gain adjustment control part A, B, C, D Deflection amount management point

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Tadayoshi Katsuyoshi 1-3-2 Shibaura, Minato-ku, Tokyo Shimizu Corporation (72) Inventor Kuniaki Nakahara 1-2-3 Shibaura, Minato-ku, Tokyo (56) References JP-A-3-281825 (JP, A) JP-A-2-108790 (JP, A) JP-A-4-198588 (JP, A) JP-A-5-280067 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) E02F 5/02 E02D 17/13 E21B 7/10 E21B 44/00

Claims (3)

(57) [Claims] 1. A method for automatically correcting a deviation from a control target of an excavator when excavating ground by lowering the excavator, the method comprising: controlling a position of the excavator during excavation; Calculating the amount of deviation with respect to and the rate of change of the amount of deviation, and performing fuzzy inference on the amount of deviation and the rate of change of the amount of deviation to be provided on the excavator to expand and contract. A displacement correction control process that controls the expansion and contraction of the jack portion that brings the excavator closer to the target by a predetermined short period, and an average value of the displacement amount, and fuzzy inference on the sign reversal frequency of the displacement amount. A gain adjustment control step of controlling the expansion and contraction of the jack portion for each long cycle longer than the short cycle. 2. An apparatus for automatically correcting a deviation from a control target of the excavator when excavating the ground by lowering the excavator, the control target of a position of the excavator during excavation. A deviation calculation amount, and a measurement calculation processing unit that calculates a rate of change of the deviation amount; and a fuzzy inference is performed on the deviation amount and the change rate of the deviation amount, and the excavator is provided with: A displacement correction control unit that controls the expansion and contraction of the jack unit that brings the excavator closer to the target by expanding and contracting, every predetermined short period, and an average value of the deviation amount, and a sign reversal frequency of the deviation amount. An automatic excursion correction device for an excavator, comprising: a gain adjustment control unit that performs fuzzy inference and controls expansion and contraction of the jack unit for each long cycle longer than the short cycle. 3. The excavator according to claim 1, wherein the excavator performs the excavation while correcting the excursion of the excavator. An upper limit and a lower limit are set for the excavator, and when the excavator is deviated within a control value between the upper limit and the lower limit, the excavating speed of the excavator is increased, and An excavating speed control method for an excavator, wherein the excavating speed of the excavator is reduced when the excavator is displaced out of a control value.