JPH07166797A - Method for grasping existing segment form - Google Patents

Abstract

PURPOSE:To easily and correctly grasp the assembling form of an existing segment by moving a sensor in the circumferential direction, obtaining the coordinates from a turning angle and the distance between the turning axis and the inner peripheral surface, and performing coordinate transformation to make an elliptical display. CONSTITUTION:Pre-sensing sensors are disposed in the circumferential direction centering round the turning axis in the horizontal direction in the state of separating from a boring machine 1 in the rear of a shield machine 2 in the existing segments 3, 3. Subsequently, each sensor is stopped at plural positions opposite to the inner peripheral surface of the segment 3 to measure the distance from the sensor to the inner peripheral surface of the segment 3. The turning angle of the sensor and the distance from the turning axis to the segment 3 are calculated to obtain the coordinates on the inner peripheral surface on the segment 3. Subsequently, the center of gravity of the center point group is calculated to compute the amount of eccentricity, and coordinate-transformed to create an approximate ellipse from the coordinate point group. The ellipse is displayed. Thus, an assembling error due to deformation of the segment 3 is avoided to shorten the term of works.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for grasping the shape of an existing segment for accurately grasping the shape of a segment used in a shield excavation method and already assembled.

[0002]

2. Description of the Related Art As is well known, there is a shield construction method as a method of constructing a tunnel. As one of the construction methods of this type, a shield excavator is used to advance the ground while excavating the natural ground. There is a method of assembling segments in a ring shape to obtain a progressive reaction force on the rear side.

In the segment assembling apparatus used for such a shield construction method, the positioning of the segment is performed by the scanning control. That is, when assembling the segments, press the concavo-convex guides provided between the abutted segments while assembling, and when connecting the abutted segments, if the bolts do not easily enter the bolt holes, intermittently use the bolts. The segment was moved little by little while moving forward and backward, and the bolt was inserted into the bolt hole when the bolt holes were aligned.

[0004]

However, in such a conventional segment assembling apparatus, the shape of the segment is equal in the normal direction to the entire circumference of the segment to which disturbance such as earth pressure is attached. Since it does not work, even if the segment is assembled in a shape close to a perfect circle, the segment may be deformed to some extent after the assembly. If the deformation of such a segment is left as it is, the roundness of the existing segment will serve as a position reference when assembling a new segment, so errors due to deformation will accumulate and time will be required in the assembling work by the copying control. Not only is there a problem that it will take time, but if the accumulated error becomes large, there is a fear that it will eventually become impossible to assemble.

The present invention has been made in view of such conventional problems, and an object thereof is to easily and accurately determine the shape of an existing segment which serves as a positioning reference when a new segment is assembled. It is to provide a method of grasping the shape of an existing segment that can be grasped.

[0006]

In order to achieve the above object, the present invention is a method for grasping the shape of an existing segment that is annularly assembled to the rear end of a shield excavator as it excavates. Installed in the existing segment in a state of being separated from the shield machine in the rear of the shield machine, using a sensor that swings in the circumferential direction about a horizontal swing axis that substantially matches the central axis of the segment, This sensor is stopped at a plurality of positions so as to face the inner circumferential surface of the segment, and the distance from the sensor to the inner circumferential surface of the segment is measured. The turning angle of the sensor and the turning axis of the segment The distance to the peripheral surface is calculated, and the coordinates (X, Y) of a plurality of positions on the inner peripheral surface of the segment are calculated based on the calculation result. A plurality of arbitrary three points selected from the mark are taken out to obtain a plurality of center point groups, the center of gravity of these center point groups is calculated, and the eccentricity amount (Δx, Δy) between the segment center and the sensor rotation center is calculated. The coordinates (X,
Y) is coordinate-converted based on this eccentricity to obtain a coordinate point group (X1, Y1), and this coordinate point group (X1, Y1)
The feature is that the existing segment shape is grasped by creating an ellipse that is approximated based on 1) and displaying this ellipse.

[0007]

With the above structure, the shape of the existing segment, which serves as a positioning reference when a new segment is assembled, can be easily and accurately grasped by making an elliptic approximation.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of the present invention will be described in detail below with reference to the accompanying drawings. FIG. 1 shows a segment assembling apparatus of a shield construction method to which the present invention is applied. The assembling apparatus 1 shown in FIG. 1 has shields inside segments 3 and 3 which are assembled behind a shield machine 2 in a hollow cylindrical shape. It is installed separately from the excavator 2.

The segment assembling apparatus 1 has four groups for supporting the inner peripheral surface of the existing segment 3 by supporting it and moving the entire apparatus forward with the assembling work of a new segment. Support device 4 and this support device 4
A support beam 5 supported in the horizontal direction and a substantially hexagonal frame shape that can swing about a horizontal axis that is supported at the tip of the support beam 5 and is orthogonal to the axis of the tunnel constructed by the existing segments 3 and 3. Of the swing frame 6, an annular swing frame 7 supported by the front portion of the swing frame 6 and capable of swinging about a swing axis orthogonal to the swing surface of the swing frame 6, and the swing frame 7 The eclectic device 9 is configured to be movable forward and backward in the radial direction of the revolving frame 7 with respect to a pair of repelling arms 8 and 8 protruding from the front part of the revolving frame 7.

A segment supply device 10 for supplying a new segment 3 to a lower portion of the erector device 9 is arranged below the segment assembly device 1 so as to pass below the support device 4.

The support device 4 comprises a first fixed holding device 11, a first reaction force support device 12, and a second device in order from the rear of the tunnel.
The reaction force support device 13 and the second fixed holding device 14 of FIG. The first and second fixed holding devices 11 and 13 are bridged through a pair of fixed support legs 15 arranged on both left and right sides of the tunnel and extending in a vertical direction, and an upper jack 16 from an upper portion of the fixed support legs 15. A crescent-shaped upper fixed shoe 17 that can be separated from and contacted with the upper portion of the inner peripheral surface of the existing segment 3 by the vertical movement of the jack 16 and the vertical movement from the lower part of the fixed support leg 15 through the lower jack 18. The lower fixed shoe 19 which can be separated from the lower portion of the inner peripheral surface of the existing segment 3 and the side of the inner peripheral surface of the existing segment 3 can be separated from the side portion of the fixed support leg 15 by moving forward and backward. A side fixing shoe 20 is provided.

On the other hand, the first and second reaction force supporting devices 12,
Reference numeral 14 denotes a pair of support legs 21 arranged on both the left and right sides of the tunnel and extending in the vertical direction, and bridged from the upper part of the support legs 21 via an upper jack 22. Crescent-shaped upper shoe 23 that can be attached to and detached from the upper portion of the inner peripheral surface, and a lower portion that can be separated from the lower portion of the inner peripheral surface of the existing segment 3 by moving up and down through the lower jack 24 from the lower portion of the support leg 21. And a shoe 25. Then, the fixed holding device 11, 1
Between the 3 and the reaction force supporting devices 12 and 14, a jack (not shown) is provided for advancing the fixed holding devices 11 and 13 by applying a reaction force to the reaction force supporting devices 12 and 14.

The support beam 5 is fixed to the first and second fixed holding devices 11 and 13, and is movable only in the front-rear direction with respect to the first and second reaction force supporting devices 12 and 14. Supported by.

As shown in FIGS. 2 and 3, the erector unit 9 is segmented by a pair of lifting jacks 26, 26 provided at the tips of the rebound arms 8, 8 so that it can be moved forward and backward in the same figure. 3 includes a crescent-shaped erector device body 27 that matches the inner surface shape of 3, and a support mechanism 28 that supports the segment 3 supplied by the segment supply device 10 to the erector device body 27, and this support mechanism 2
8. A plurality of presensing sensors S arranged around 8
1, S2, S4 (three in this embodiment) and the like.

The support mechanism 28 is fixed to the jack 29 extending in the radial direction of the revolving frame 7 from the eclectic device body 27, and is fixed to the plunger 30 of the jack 29 to move up and down in the fixed cylinder 31 of the erector device body 27 in the vertical direction. A movable barrel 32, a chuck 33 protruding from the movable barrel 32 to the outside in the radial direction, and a drive mechanism 3 provided in the movable barrel 32 for rotating the chuck 33 about its axis.
4 and.

A 1/4 ring-shaped convex portion 35 is fixedly provided at the tip of the chuck 33, and the jack 29 side of the convex portion 35 is inclined inward toward the jack 29 side as shown in FIG. A surface 36 is formed. On the other hand, a hole 37 is formed in the center of the segment 3 and the chuck 3 is placed in the hole 37.
A cylindrical metal fitting 38 with which 3 is engaged is fitted, and an inclined surface 39 that engages with the inclined surface 36 of the chuck 33 is formed on the cylindrical metal fitting 38. The engagement between the chuck 33 and the cylindrical metal fitting 38 allows the chuck 33 to be inserted into the cylindrical metal fitting 38 when the chuck 33 is set at a predetermined rotation angle, and the chuck 33 is rotated, for example, by 90 degrees in the inserted state. At this time, the segment 3 is supported by the chuck 33 by the surface contact between the inclined surfaces 36 and 39.

Further, the presensing sensors S1, S2,
As shown in FIG. 2, S4 is arranged on the outer peripheral surface of the erector device body 27 along the circumferential direction with a span of, for example, about 40 to 140 mm. Pre-sensing sensor S1
Is located outside the erector device body 27, and the presensing sensor S2 is located near the center line of the erector device body 27. Then, in a state in which the support mechanism 28 of the erector device 9 does not support the segment 3, as described later, the revolving frame 7 is revolved and the lifting jack 2 is moved.
6 is driven to position the elector device main body 27 so as to face the inner circumferential surfaces of the rings of the existing segments 3 and 3, and the laser light is emitted and the reflected light is received to measure the respective distances. Sensor.

The segment assembling apparatus 1 having the above construction assembles a new segment 3 through the following steps. That is, after the assembling of one ring (in this embodiment, one ring is made up of seven segments) is completed and before the excavation is completed and before the next segment of one ring is newly assembled. , Presensing sensor S
1, S2 is used to measure the positions of the existing segments 3 and 3 to grasp the mounting shape, and after calculating the distortion amount (correction value) with the perfect circle for each segment or each group,
A new segment 3 is assembled based on this correction value. The presensed data is learned, and even if the cross-sectional shape of the segments 3 and 3 changes for each ring, the shape of the new segment 3 approximates a perfect circle. Such an operation,
The analysis is performed by the control means (not shown) provided in the segment assembling apparatus 1.

Next, a method of measuring the position of the existing segment, a method of grasping the shape of the existing segment, and a method of assembling the segment, which are performed by the control means, will be described.

[Presensing] First, the turning frame 7
The inner peripheral surface coordinates (X, Y) of the existing segment (segment for one ring assembled immediately before the segment to be newly assembled) 3, 3 when the turning axis of (3) is set as the center (0, 0) Obtain at multiple locations. In the state where the support mechanism 28 of the erector device 9 does not support the segment 3, the revolving frame 7 is revolved around its revolving axis (the oscillating frame 6 is fixed), and the erector device main body 27. At a plurality of positions (in this embodiment, 180 degrees, 90
(4 places of 0 degree, 0 degree, and -90 degrees), and each of the pre-sensing sensors S1 and S2 from the existing segment 3,
This is done by collecting distance data to the inner peripheral surface of No. 3. The turning angles of the sensors S1 and S2 are calculated as follows. That is, as shown in FIG. 5, the turning angle of the presensing sensor S1 is the erector turning angle + θ1 + 360 * (Y−Yc) / (2π * Y
r), that is, the erector turning angle + θ1 + (Y-Yc)
* 180 / (π * Yr), the turning angle of the presensing sensor S2 is the same as the erector turning angle + θ2 + (Y-
Yc) * 180 / (π * Yr), the turning angle of the pre-sensing sensor S4 is the erector turning angle + θ4 + (Y−Y
c) * 180 / (π * Yr). Here, Y is the swing stroke (mm) on the arc when sensing, Yc
Is a swing stroke (mm) (fixed value) of the support mechanism 28 when the center of the erector is mechanically centered, and Yr is a swing turning radius (mm).

Here, the presensing sensors S1 and S
Since S2 and S4 are arranged on an arc of the erector device main body 27, it is necessary to correct the height position output value of the sensor. The sensor S2 is used as a parameter for correcting the sensor S1 because it does not need to be corrected because the raising / lowering direction of the raising / lowering jack 26 substantially coincides with the radial direction of the revolving frame 7. That is, the correction amount ΔH for the height position of the sensor S1 is ΔH = r− (Δr * cos θ1 + (r−Δr) * cos
θa), the correction amount ΔH for the height position of the sensor S4 is ΔH = r− (Δr * cos θ4 + (r−Δr) * cos
θa) It is calculated by the equation. Note that θ1 is the mounting angle of the sensor S1 with respect to the center position of the erector device body 27, r is the distance (mm) from the turning center to the sensor S1, Δr is (output value of the sensor S1) −H0 (mm), and θ1 is asin.
((Δr * sin θ1) / (r−Δr)). Here, H0 is a value mechanically designed to pass through the turning center when the output value is H0.

The distance from the turning axis of the turning frame 7 to the inner peripheral surfaces of the existing segments 3 and 3 is calculated as follows. That is, the separation distance of the pre-sensing sensor S1 (the length from the actual turning center of the erector to the surface of the existing segment) is the separation distance (S1) = ((elevation stroke (left) + elevation stroke (right)) / 2 + Lc + sensor S1 output value + Δ
The separation distance of the pre-sensing sensor S2 is calculated by the formula H, and the separation distance (S2): ((up / down stroke (left) + up / down stroke (right)) / 2 + Lc + sensor S2 output value expression. , Since the distance is measured only by the lifting stroke, the correction value (m
m). Here, the reason why the left and right elevating strokes are averaged is that the respective jacks are servo-controlled, and the allowable error thereof is taken into consideration.

If the turning angles of the presensing sensors S1, S2, S4 stopped at a plurality of points and the distance from the turning axis to the inner peripheral surfaces of the segments 3 and 3 are calculated as described above, the segment 3 is calculated. , 3 at a plurality of positions on the inner peripheral surface can be obtained.

[Analysis of cross-sectional shape of existing segment] From the collected plural coordinate data, three data which are not adjacent to each other are selected, and the coordinate point group (X, Y) is inputted to the control means for each data. It Assuming that the point cloud (X, Y) consists of points on the circumference, the intersection of the vertical bisectors of three straight lines formed by any three points in the point cloud passes through the center of the circle. Therefore, the center point group can be obtained by extracting any three points from the point group (X, Y) and finding the intersections thereof. By repeating the same operation, a plurality of such ellipses are created and the center point group of the ellipses is obtained for each of them.

Next, the center of gravity method calculation is performed on the obtained center point group, and the arithmetic mean thereof is defined as (Δx, Δy). In this case, it is preferable to take out three points so that a combination having adjacent points is excluded in consideration of a point that an error may increase when neighboring points are used as arbitrary three points. More center of gravity calculation is performed, and the obtained point is set as the center of the actually assembled segments 3 and 3.

After that, the collected data is converted into data in a new coordinate system centered on the segment center, and the eccentricity (Δx, Δ) between the segment center and the turning center of the sensor is converted.
y) is calculated. Next, using this (Δx, Δy),
The coordinates of (X, Y) are converted by the following equation, and the point group (X1,
Y1) is created. (X1i, y1i) = (xi−Δx, yi−Δy) In this case, the tilt angle θ of the ellipse, the X-axis radius (θ direction) α when the ellipse is approximated, and the Y-axis radius (θ) when the ellipse is approximated.
In the (+ π / 2 direction) β, θ is assumed to be 0 degree, and the point group (X1, Y1) is rotated by −θ by the following equation, and further the point group (X2
, Y2) is created. cos−θ−sin−θ (x2i, y2i) = (sin−θ, cos−θ) (x1i,
y1i) Next, for the point cloud (X2, Y2), for example, a minimum of 2
The following formula is approximated by multiplication. Y2 ² = β ² -X ² * α ² / β ² Also, the absolute value sum ΣΔr of the difference between the distance from the origin and the elliptic radius with respect to the point group (X2, Y2) is calculated and stored in the table. Prior to this, the distance from the origin is rd, the radius at the azimuth angle θ0 on the ellipse is ro, the x coordinate at the azimuth angle θo on the ellipse is x0, and the y coordinate at the azimuth angle θo on the ellipse is y.
0, where θ0 is the angle of the straight line formed by the origin and the point, rd = sqrt (x2i ² + y2i ² ) ro = sqrt (x0 ² + y0 ² ) xo = 1 / ((1 / α) ² + (tan (θ0 ) /
β) ² ) yo = sqrt (β ² − (β * x0 / α) ² ) θo = atan (y2i / x2i) are calculated.

At this time, if θ is 90 degrees or less, 0 to 90
The coordinate axes are rotated at intervals of 0.5 degree between each degree, and the ellipse formula is approximated by using the least squares method on each coordinate axis (180 data), and the absolute value sum of the respective errors is calculated and stored in the table. store. Then, ΣΔr in the table is compared, and an ellipse having θ when the sum of absolute values ΣΔr is the smallest approximates the shape of the existing segment ring immediately before a new segment is assembled. In this way, the existing segment shape can be grasped by elliptic approximation of the coordinate point group obtained by presensing. The ellipse obtained as described above is displayed on the display as needed.

The absolute value Δri of the difference between the distance from the origin and the ellipse radius for the point group (X2, Y2) is calculated and stored in the table.

Next, it is judged whether or not the condition of the following equation is satisfied. If not satisfied, the maximum Δri is obtained (x2
i, y2i) is deleted and the process returns to the input of the coordinate point group (X, Y). Δri <ΣΔri / n + ε sqrt (Σ (Δri ² ) / (n−2)) <Rs Here, n is the effective score.

If the number of measurement points is less than or equal to the minimum number of measurement points, it is assumed that proper approximation could not be made.

[Calculation of Distortion Amount] In this embodiment, as shown in FIG. 6, seven segments A1, which form one ring,
A2, A3, A4, B1, B2, K are group 1
(G1): Segments A1, A2, Group 2 (G
2): Segments A3, A4, Group 3 (G3):
Segments B1, B2, and K are divided into three groups, and a value obtained by integrating the radius difference between the true circle and the segment ring for each group is the distortion amount. Since this amount of distortion means a difference in radius from the true circle, it may be reduced to bring the shape closer to the true circle. Therefore, the correction amount of the lifting stroke for the desired ring shape matches the value obtained by multiplying the obtained distortion amount by minus (-). Here, the distortion amount ΣΔri in a certain group i is

[Equation 1] It is calculated by the formula. Note that si is the start angle of group i (changes depending on right grouping / left grouping), ei is the end angle of group i (changes depending on right grouping / left grouping), and rθ is the radius at the azimuth angle θ on the approximate ellipse. , Α is the x-axis radius of the approximate ellipse, β
Is the y-axis radius of the approximate ellipse.

[Correction value calculation of existing segment] Two patterns (CASE1: [G1 +
G3] VSG2, CASE2: G3 vs G2) is set so that it can be selected. The correction value may be calculated for each segment, or may be calculated for each group. In this embodiment, a method of calculating a correction value for each group is adopted. It should be noted that group 1 does not require correction because it forms the basis of the ring formed by the segments. However, the total sum of the correction values in one ring needs to be approximately ± 0. This is because, for example, if the correction value for group 1 is set to +3 mm, errors in the turning direction accumulate and the upper segment cannot be assembled.

[Segment Support and Positioning] The segment 3 is conveyed to the predetermined position by the segment feeder 10 and positioned. Since the assembling apparatus 1 accurately supports the segment 3, the position of the segment 3 is grasped by a laser sensor (not shown). When the position of the segment 3 is accurately grasped, the support mechanism 28 of the elector device 9 is
The segment 3 is supported by. The segment 3 supported by the erector device 9 adjusts the inclination of the revolving surface of the revolving frame 7 by revolving the revolving frame 6, and also opposes the assembly position of the segment 3 by revolving the revolving frame 7. It is positioned at the position where The segment 3 positioned at a position facing the mounting position is set to the mounting position by driving the lifting jack 26. The lifting stroke of the lifting jack 26 at this time is corrected by an amount corresponding to the above-described correction amount, and the segment 3 is positioned with the corrected stroke.

[Segment fixation] When the positioning of the segment 3 is completed, the new segment 3
Bolts (not shown) and bolt holes of the existing segment are inserted, and nuts are fastened. When the bolts are inserted into all the bolt holes in one segment, the tightening torque of the bolts is adjusted, and the assembly of one segment is completed. [Advancing movement of the assembling device] After the above assembling and fixing work is repeated and one ring segment is assembled and fixed,
When a space for one ring is formed at the rear end by excavation of the shield machine 2 in which the excavation work is continued during this fixing work, the first and second fixing holding devices 11, 13 are formed.
The side fixed shoe 20 and the upper and lower fixed shoes 1
The first and second reaction force supporting devices 1 are made to degenerate 7 and 19.
After moving forward while applying a reaction force to the Nos. 2 and 14, the upper and lower fixing shoes 17 and 19 are extended and closely fixed to the inner peripheral surfaces of the existing segments 3 and 3. Next, after retracting the upper and lower shoes 23, 25 of the first and second reaction force supporting devices 12, 14 and moving them forward, the upper and lower shoes 23, 25 are extended to extend the inner circumference of the existing segments 3, 3. Fix closely to the surface. The side fixing shoes 20 of the first and second fixed holding devices 11 and 13 may be extended thereafter, or may be extended immediately after the fixed holding devices 11 and 13 are advanced. Of course, after the shield work is completed, or when it is diverted to another shield work, if the supporting device 4 retracts the entire device, the assembling device 1 can be moved to the shield vertical hole and pulled up to the ground. .

[Learning] As described above, the correction value is calculated from the presensing result to assemble a segment for one ring, and after the excavation is completed, the segment just assembled is presensed. Assembly device 1
The control means of (1) evaluates how the shape of the assembled segment in consideration of the correction value is deformed as compared with the previously presensed segment, and accumulates it as learning data.

For this reason, the control means has a learning system that learns for each group what correction amount should be applied to cause a change in a certain distortion amount based on the next input / output. Have The learning system gave it to the previous ring when the change amount of the strain amount in the previously analyzed ring and the strain amount in the ring analyzed this time (the ring immediately before the ring to be newly assembled) was input. The correction amount is output.

As shown in FIG. 7, each group has a matrix of a predetermined membership function (MF) showing the correlation between the strain change amount dr (horizontal axis) and the matching degree (vertical axis). . The MF has a trapezoidal or triangular shape, and each time new data is input, it is statistically analyzed and its shape is updated. Here, the amount of distortion change is calculated for each group, and assuming that the correction given to a certain group may change the amount of distortion of another group, the total change for each group is It is supposed to be examined.

The distortion correction amount is applied to the updated matrix of MFs, and the degree of agreement is obtained for each MF. Three MFs are arranged per matrix row, and the matching degree of each row is calculated by multiplying the matching degrees of each. Among the obtained matching degrees, those which exceed a predetermined threshold value are selected, and the assigned correction values are weighted and added according to the respective matching degrees. The correction value is determined by rounding the correction value thus obtained at a pitch of 0.5 mm, for example.

In reality, the matrix of MFs is arranged on the network. If the change in the amount of distortion is significantly different even with the same correction value during learning, it is determined that the population is different and a new synapse (row of matrix) is automatically expanded. Further, various weighting calculations and threshold calculation for determining the above-mentioned correction value are also processed on this network.

By the way, there is a through-out as the simplest procedure for calculating the control amount. The control amount TOi for the group i due to the through-out is obtained by the formula TOi = ΣΔri. However, this control amount must not exceed the maximum / minimum value given by a constant. However, the control amount obtained by the through-out does not take into consideration the disturbance such as gravity, and the present distortion cannot always be eliminated. Therefore, control is performed by the learning system that uses the change due to the actual control amount as the learning data.

Since the input and output at the time of learning have a relationship of "input generated by giving output", learning this relationship results in "output required to generate input".
Is expected to be able to reason. On the other hand, a certain number of learnings are required to realize this.

However, since the learning system cannot perform inference in the initial state, the neutral value (0) is always used as the inference result, and the output during learning is always the neutral value, so that no learning can be performed.
If the inference could not be performed (low confidence), the through-out is set as the inference result. As the certainty factor, the degree of agreement such as inference is used, but if the degree of agreement is less than or equal to the maximum value of the difference / 2, it is determined that inference cannot be made, and a through-out is returned. As a result of a simple experiment, the certainty factor becomes stable from the time when 10 rings are assembled, and it is considered that the condition of this case will be satisfied.

The input at the time of inference is "desired distortion amount change", and it is desirable that the distortion amount be 0. Therefore, the current distortion amount is input and the correction amount to be given this time is output. Also, the coincidence is calculated at the same time as the inference, but this has the same meaning as the certainty.

In the above embodiment, the case where the present invention is applied to the rear independent segment assembling apparatus is shown. However, the present invention is not limited to the rear independent type and is applied to a machine mounting type, and the same operational effect is obtained. It goes without saying that you will play.

[0045]

As described above in detail with reference to the embodiments, the turning angle of the sensor and the distance from the turning axis to the inner peripheral surface of the segment are calculated, and the inner peripheral surface of the segment is calculated based on the calculation result. Coordinates (X,
Y) is calculated, the eccentricity amount (Δx, Δy) between the center of the segment and the turning center of the sensor is calculated from a plurality of coordinates, and the coordinates (X, Y) are coordinate-converted based on the eccentricity amount. An ellipse that is approximated based on the group (X1, Y1) is created, and the shape of the existing segment is grasped by displaying this ellipse. Therefore, the shape of the existing segment that serves as the positioning reference when assembling a new segment is set. The measurement can be performed easily and accurately, so that there is no assembly error due to deformation, the construction period can be shortened, and the assembly is not impossible.

[Brief description of drawings]

FIG. 1 is an overall configuration diagram of a segment assembling apparatus according to the present invention.

FIG. 2 is a front view showing details of the same elector device.

3 is a cross-sectional view of the main parts of FIG.

4 is a cross-sectional view taken along the line AA of FIG.

FIG. 5 is an explanatory diagram of presensing.

FIG. 6 is a diagram showing grouping of one ring to be assembled.

FIG. 7 is a diagram showing a matrix of membership functions used in a learning system.

[Explanation of symbols]

 1 segment assembly device 2 shield machine 3 segment 4 support device 5 support beam 6 rocking frame 7 swing frame 9 erector device 10 segment supply device 26 lift jack 27 erector device body 28 support mechanism S1, S2 presensing sensor

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Hiroyuki Morino 2-3 Kandaji-cho, Chiyoda-ku, Tokyo Obayashi Corporation Tokyo Head Office

Claims (1)

[Claims] 1. A method for grasping the shape of an existing segment that is annularly assembled to the rear end of the shield excavator when the shield machine is excavated, wherein the segment is installed in the existing segment and substantially coincides with the central axis of the segment. Using a sensor that rotates in the circumferential direction around the horizontal turning axis, the sensor is stopped at a plurality of positions so as to face the inner circumferential surface of the segment, and the sensor to the inner circumferential surface of the segment is stopped. The distance is measured, the turning angle of the sensor and the distance from the turning axis to the inner peripheral surface of the segment are calculated, and the coordinates (X, Y) of a plurality of positions on the inner peripheral surface of the segment are calculated based on the calculation result. ), And then any 3 selected from the collected coordinates.
A plurality of groups of points are taken out to obtain a plurality of center point groups, the center of gravity of these center point groups is calculated, the eccentricity amount (Δx, Δy) between the segment center and the turning center of the sensor is calculated, and the coordinates (X, Y) are calculated. ) Is coordinate-converted based on this eccentricity to obtain a coordinate point group (X1, Y1), and this coordinate point group (X1
1. A method of grasping an existing segment shape, which is characterized by creating an ellipse approximated based on Y1) and displaying the ellipse to grasp the existing segment shape.