CN108801171A - A kind of tunnel cross-section deformation analytical method and device - Google Patents

Abstract

The invention discloses a tunnel section deformation analysis method, the method comprising: collecting tunnel section information to obtain a point cloud data set; analyzing and comparing the point cloud data sets collected in different periods of the tunnel section to be detected to determine the tunnel to be detected The deformation trend of the cross section. As a result, the section information of the tunnel in different periods can be visualized, the deformation of the section can be determined through the comparison of the point cloud data sets, and the trend of the tunnel section over time can be detected overall and comprehensively, and the measurement accuracy of the application is high. The measurement method is relatively simple and easy to implement.

Description

A method and device for analyzing deformation of a tunnel section

technical field

The invention relates to the technical field of tunnel detection, in particular to a tunnel section deformation analysis method and device.

Background technique

With the continuous improvement of the level of urbanization, the urban population has also increased rapidly, causing a series of problems such as traffic congestion and environmental pollution. As a tool to alleviate urban traffic congestion, urban subway has developed rapidly.

Since urban subway lines generally pass through main arterial roads and densely populated central areas, deformation of the underground tunnel itself, pipelines and surrounding buildings will be caused during the construction of the subway. At the same time, during the operation of the subway, the vertical displacement, horizontal displacement, cracks and other deformations caused by the nature of the soil itself, groundwater, and the load of the ground buildings on the tunnel may be obvious in some sections. Deformation monitoring and analysis of monitoring data will cause unimaginably serious consequences.

The formulation and implementation of a deformation monitoring plan is a prerequisite for deformation prediction, an important link in the entire process of informatization, and has a major impact on deformation prediction. A reasonable monitoring plan is an important part of the normal and orderly construction of subway construction. Conditions, but also provide experience for the construction of similar projects, to avoid risks and accidents. Due to the large time span, complex influencing factors, and large social impact of disasters in the subway operation stage, the deformation monitoring of subway construction must be long-term and continuous.

Among them, the section deformation monitoring of the urban subway subway is a very important and complex system engineering. The design of the monitoring scheme is the basic content of the entire monitoring process. The monitoring content mainly includes the vertical settlement of the tunnel, the horizontal displacement and the convergence deformation of the section. A single data index is difficult to reflect the overall tunnel deformation. At present, most methods use a total station to select a limited number of discrete points on the section for measurement. Although the measurement accuracy of this total station is high, it is difficult for the measured discrete points to reflect the overall deformation of the tunnel section. The analysis and processing of monitoring data and the analysis method of deformation data are still not perfect.

At present, no effective solution has been proposed for the problem that the existing technology cannot simultaneously reflect the overall and local changes of the tunnel section and the deformation trend of the tunnel section over time.

Contents of the invention

Embodiments of the present invention provide a tunnel section deformation analysis method and device, which can solve the problem in the prior art that the overall and local changes of the tunnel section and the deformation trend of the tunnel section over time cannot be reflected at the same time.

In order to solve the above technical problems, in the first aspect, an embodiment of the present invention provides a method for analyzing deformation of a tunnel section, the method comprising:

Step 1, collect tunnel section information to obtain point cloud datasets;

Step 2, analyzing and comparing the point cloud data sets collected at different periods of the tunnel section to be detected, so as to determine the deformation trend of the tunnel section to be detected.

Further, step 1 includes:

Step 11, collect tunnel section information by scanning the same tunnel section to be detected in different periods with a 3D laser scanner; acquire point cloud data sets P ₁ and P ₂ in two periods according to the tunnel section information.

Further, step 2 includes:

Step 21, respectively performing circle fitting on the point cloud datasets of the two periods in their respective coordinate systems, so as to determine the circle centers of the two point cloud datasets respectively;

Step 22, placing the two point cloud datasets in the same coordinate system, and then aligning the centers of the circles to perform the first registration of the two point cloud datasets;

Step 23, performing a second registration on the two point cloud datasets after the first registration according to the ICP algorithm;

Step 24, randomly select one of the point cloud data sets as the reference point cloud data set, and use the ANN algorithm to determine the corresponding points of each reference point in the other point cloud data set;

Step 25, analyzing the relationship between each pair of reference points and corresponding points to determine the deformation trend of the tunnel section to be detected;

Wherein, the reference point is a point in the reference point cloud data set.

Further, the step 21 specifically includes: performing circle fitting on the point cloud data sets of the two periods in their respective coordinate systems according to the least square method.

Further, step 23 includes:

Step 231, calculate the corresponding neighboring points of each point in the point cloud data set P ₂ in the point cloud data set P ₁ , to obtain the first group of N corresponding point pairs;

Step 232, determine the distance and the minimum translation parameter of the first group of N corresponding point pairs and rotation parameters Among them, the translation parameter and rotation parameters is the parameter in the rigid body transformation;

Step 233, when the distance sum of the first group of N corresponding point pairs is greater than or equal to the preset distance, according to the point cloud data set P ₂ , the translation parameter and rotation parameters Determine the point cloud data set _P'2 ; calculate the corresponding adjacent points of each point in the point cloud data set _P'2 in the point cloud data set P1, to obtain the _second group of N corresponding point pairs;

Step 234, when the distance sum of the second group of N corresponding point pairs is greater than or equal to the preset distance, re-determine a new point cloud data set and perform iterative calculations until the distance of the i-th group of corresponding point pairs is less than the preset distance Set distance.

Further, the step 231 specifically includes: according to preset conditions, and using the KD-tree structure to determine the corresponding neighboring points of each point of P ₂ in P ₁ to form a set of corresponding point pairs {(P _{1, i} , P _2,i |i=1,2,...,N)};

Among them, the preset condition is: ε(P ₁ , P ₂ )=min(d ² (P _1,i P _2,i )); P _1,i and P _2,i are corresponding point pairs of the i-th group, P _{1, i} belongs to P ₁ , P _{2, i} belongs to P ₂ .

Further, the step 232 specifically includes: step 2321, substituting N corresponding point pairs into the objective function; determining the rigid body transformation matrix through an iterative algorithm To determine the distance of the corresponding point pair and the minimum translation parameter and rotation parameters

Among them, the objective function value is the distance sum of N corresponding point pairs, and the formula is:

Among them, N is the number of points in P ₁ , P _1,i and P _2,i are point pairs corresponding to group i, P _1,i belongs to P ₁ , P _2,i belongs to P ₂ . is a 3×1 translation matrix, is a 3×3 rotation matrix.

Further, the objective function value is the sum of the distances of N corresponding point pairs, and the formula is:

Among them, N is the number of points in P ₁ , P _1,i and P _2,i are point pairs corresponding to group i, P _1,i belongs to P ₁ , P _2,i belongs to P ₂ . is a 3×1 translation matrix, is a 3×3 rotation matrix, and w _i is a constraint weight factor.

further,

Among them, DDF _k (p) is the deviation factor of point p, is the regularized standard deviation of point p.

Further, the step 231 further includes: deleting corresponding point pairs whose distance is greater than a distance threshold.

Further, the distance of the corresponding point pair is the projection distance of the Euclidean distance between two points on the extension line of the radius of the coincident circle center.

In the second aspect, an embodiment of the present invention provides a tunnel section deformation analysis device, the device is applied in the method described in the first aspect, and the device includes:

The collection unit is used to collect tunnel section information to obtain point cloud data sets;

The comparative analysis unit is used to analyze and compare the point cloud data sets collected in different periods of the tunnel section to be detected, so as to determine the deformation trend of the tunnel section to be detected.

Further, the acquisition unit is also used to collect tunnel section information by scanning the same tunnel section to be detected in different periods with a three-dimensional laser scanner; acquire point cloud data sets in two periods according to the tunnel section information P ₁ and P ₂ .

Further, the comparative analysis unit is also used to perform circle fitting on the point cloud datasets of the two periods in their respective coordinate systems, so as to respectively determine the centers of the two point cloud datasets; The data sets are placed in the same coordinate system, and then the centers of the circles are coincident to perform the first registration of the two point cloud data sets; according to the ICP algorithm, the second registration is performed on the two point cloud data sets after the first registration; Randomly select one of the point cloud datasets as the benchmark point cloud dataset, and use the ANN algorithm to determine the corresponding points of each benchmark point in the other point cloud dataset; analyze the relationship between each pair of benchmark points and corresponding points, to determine the deformation trend of the tunnel section to be detected; wherein, the reference point is a point in the reference point cloud data set.

By applying the technical solution of the present invention, the section information of the tunnel in different periods can be displayed visually, and the deformation of the section can be determined through the comparison of the point cloud data sets, so as to realize the overall and comprehensive detection of the trend of the tunnel section over time. The applied measurement accuracy is relatively high, and the measurement method is relatively simple and easy to implement.

Description of drawings

Fig. 1 is a flow chart of a method for analyzing deformation of a tunnel section according to an embodiment of the present invention;

Fig. 2 is a flow chart of a tunnel section deformation analysis method according to an embodiment of the present invention;

Fig. 3 is a flow chart of a tunnel section deformation analysis method according to an embodiment of the present invention;

Fig. 4 is a flow chart of a tunnel section deformation analysis method according to an embodiment of the present invention;

Fig. 5 is a schematic diagram of calculating the distance between corresponding point pairs according to an embodiment of the present invention;

6 is a schematic diagram of two point cloud datasets after the first registration according to an embodiment of the present invention;

Fig. 7 is a schematic diagram of two phases of point cloud datasets after registration in the second phase according to an embodiment of the present invention;

8 is a schematic diagram of a global deformation trend of a tunnel section to be detected according to an embodiment of the present invention;

Fig. 9 is a schematic diagram of a global deformation result calculated from a real tunnel section to be detected according to an embodiment of the present invention;

Fig. 10 is a structural block diagram of a tunnel section deformation analysis device according to an embodiment of the present invention.

Detailed ways

The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, and are not intended to limit the present invention.

In the following description, use of suffixes such as 'module', 'part' or 'unit' for denoting elements is only for facilitating description of the present invention and has no specific meaning by itself. Therefore, 'module', 'part' or 'unit' may be used in combination.

In order to solve the problem that the existing technology cannot simultaneously reflect the overall and local changes of the tunnel section and the deformation trend of the tunnel section over time, an embodiment of the present invention provides a method for analyzing the deformation of the tunnel section. The method includes:

Step 1, collect tunnel section information to obtain point cloud datasets;

Step 2: Analyze and compare the point cloud data sets collected at different periods of the tunnel section to be detected to determine the deformation trend of the tunnel section to be detected.

By applying the technical solution of the present invention, the section information of the tunnel in different periods can be visualized and displayed, and the deformation of the section can be determined by comparing the point cloud data sets, so as to realize the overall and comprehensive detection of the trend of the tunnel section over time. Moreover, the measurement accuracy of the present application is relatively high, and the measurement method is relatively simple and easy to implement.

In a possible implementation, as shown in Figure 2, step 1 includes:

Step 11, collect tunnel section information by scanning the same tunnel section to be detected in different periods with a 3D laser scanner; acquire point cloud data sets P ₁ and P ₂ in two periods according to the tunnel section information.

It is understandable that the data collected by the 3D laser scanner is more comprehensive and the collection accuracy is higher. The tunnel section to be detected can be a subway tunnel section, and the same tunnel section can be scanned at different times by the 3D laser scanner. This application uses the comparative analysis of the scanning results of two periods as an example for introduction.

In a possible implementation, as shown in Figure 3, step 2 includes:

Step 21, respectively performing circle fitting on the point cloud datasets of the two periods in their respective coordinate systems, so as to determine the circle centers of the two point cloud datasets respectively;

Wherein, step 21 specifically includes: performing circle fitting on the point cloud data sets of the two periods in their respective coordinate systems according to the least square method.

It should be noted that the fitting method is not limited to the least square method, as long as it can achieve the effect of fitting the point cloud data set into a curve.

Step 22, placing the two point cloud datasets in the same coordinate system, and aligning the centers of the circles, so as to perform the first registration on the two point cloud datasets;

Among them, the first registration can be understood as preliminary registration, and the schematic diagram of the two point cloud datasets after the first registration can refer to FIG. 6 .

Step 23, according to the ICP (Iterative Closest Point, iterative closest point algorithm) algorithm, the two point cloud data sets after the first registration are subjected to the second registration; and the schematic diagram of the two point cloud data sets after the second registration Refer to Figure 7.

Among them, the second registration is more accurate than the first registration, and the best registration position of the two point cloud datasets can be obtained in the same coordinate system by using the ICP algorithm, which can also be understood as the best registration image.

Step 24, randomly select one of the point cloud datasets as the reference point cloud dataset, and use the ANN (Approximate Nearest Neighbor) algorithm to determine the corresponding point cloud data set of each reference point in the other point cloud dataset. point;

Step 25, analyzing the relationship between each pair of reference points and corresponding points to determine the deformation trend of the tunnel section to be detected; wherein, the reference points are points in the reference point cloud data set.

Using a high-precision 3D laser scanner, ICP registration method is used to stitch and register the two phases of tunnel cross-section measurement data, and then the two phases of point clouds are connected to adjacent points, so that the cross-sectional shapes of the tunnel in different periods can be placed in the same Visual display in the coordinate system, and determine the deformation of the section through the comparison between corresponding points, can realize the overall and comprehensive detection of the deformation of the tunnel section.

In a possible implementation, as shown in Figure 4, step 23 includes:

Step 231, calculate the corresponding neighboring points of each point in the point cloud data set P ₂ in the point cloud data set P ₁ , to obtain the first group of N corresponding point pairs;

Step 231 specifically includes: according to preset conditions, and using the KD-tree structure to determine the corresponding adjacent points of each point of P ₂ in P ₁ , and form a set of corresponding point pairs {(P _{1, i} , P _{2, i} | i=1,2,...,N)};

Among them, the preset condition is: ε(P ₁ , P ₂ )=min(d ² (P _1,i P _2,i )); P _1,i and P _2,i are corresponding point pairs of the i-th group, P _{1, i} belongs to P ₁ , P _{2, i} belongs to P ₂ .

That is to say, P _1,i P _2,i that minimizes ε(P ₁ ,P ₂ ) is the corresponding point pair that is sought.

Step 231 further includes: deleting corresponding point pairs whose distance is greater than a distance threshold.

In the process of determining the corresponding point pairs, the point pairs whose distance between the point pairs is greater than the threshold value are eliminated through distance calculation, and the point pairs that do not meet the conditions can be removed, which can eliminate their influence on the subsequent calculation of the objective function, thereby improving the matching. quasi-precision.

In a possible implementation manner, the distance between the corresponding point pairs is the projection distance of the Euclidean distance between the two points on the extension line of the radius of the coincident circle center. The method of calculating the corresponding point-to-point distance can be regarded as an improvement to the ICP algorithm. On the basis of the original algorithm, when calculating the objective function, the original point-to-point error measurement method is changed to point along the radius direction The distance, that is, the error between two points is the projection distance of the Euclidean distance on the extension line of the radius. As shown in Figure 5, this method calculates the distance based on the point cloud of the tunnel section being roughly circular and its subsequent deformation characteristics, so that the optimal point cloud registration result can be obtained.

Step 232, determine the distance and the minimum translation parameter of the first group of N corresponding point pairs and rotation parameters Among them, the translation parameter and rotation parameters is the parameter in the rigid body transformation;

Wherein, step 232 specifically includes: step 2321, substituting N corresponding point pairs into the objective function; determining the rigid body transformation matrix by an iterative algorithm To determine the distance of the corresponding point pair and the minimum translation parameter and rotation parameters

Among them, the objective function value is the distance sum of N corresponding point pairs, and the formula is:

Among them, N is the number of points in P ₁ , P _1,i and P _2,i are point pairs corresponding to group i, P _1,i belongs to P ₁ , P _2,i belongs to P ₂ . is a 3×1 translation matrix, is a 3×3 rotation matrix. The objective function can also be written as an error function.

Step 233, when the distance sum of the first group of N corresponding point pairs is greater than or equal to the preset distance, according to the point cloud data set P ₂ , the translation parameter and rotation parameters Determine the point cloud data set _P'2 ; calculate the corresponding adjacent points of each point in the point cloud data set _P'2 in the point cloud data set P1, to obtain the _second group of N corresponding point pairs;

Step 234, when the distance sum of the second group of N corresponding point pairs is greater than or equal to the preset distance, re-determine a new point cloud data set and perform iterative calculation until the distance of the i-th group of corresponding point pairs is less than the preset distance .

Different from the formula of the above objective function, in a possible implementation, the objective function value is the sum of the distances of N corresponding point pairs, and the formula can be:

Among them, N is the number of points in P ₁ , P _1,i and P _2,i are point pairs corresponding to group i, P _1,i belongs to P ₁ , P _2,i belongs to P ₂ . is a 3×1 translation matrix, is a 3×3 rotation matrix, and w _i is a constraint weight factor.

Among them, DDF _k (p) is the deviation factor of point p, is the regularized standard deviation of point p. In order to eliminate the impact of outliers and noise points on the calculation of the objective function, it is necessary to assign different weights to each corresponding point pair. For example: for the corresponding point pairs judged to be noise or outliers, a smaller weight is set for them, so as to reduce the impact of abnormal corresponding point pairs on the registration accuracy and improve the registration accuracy.

The following briefly introduces the weighting factors introduced in this application.

In practical engineering applications, different regions of the registration model often have different importance. In this application, different weights are applied to the corresponding point pairs, so as to use the weights to constrain and ensure the registration accuracy of important regions of the model. In the existing research, the setting of the weight is mainly to distinguish whether the point set participates in the registration, and the weight is only 0 and 1. This method is easy to eliminate normal corresponding point pairs, which reduces the accuracy of registration. Therefore, the weight factor w _i is introduced to improve the original error function, the formula is as follows:

The constraint weight factor is defined as the importance of the point set participating in the registration, and the translation parameter can be obtained by minimizing the error function and rotation parameters The method is shown in Figure 4. In order to reduce the impact of abnormal corresponding points on the registration accuracy, the weight factor corresponding to the point cloud dataset should be inversely proportional to the size of the abnormal point measurement, that is, the greater the possibility of the point being an abnormal point, the corresponding weight of the point factor should be smaller. Therefore, the abnormal point measurement function N(p) is introduced, p is the current measurement point, and the weight factor is expressed as In order to make the metric function correctly reflect the possibility that point p is an abnormal point, the metric function is defined as:

where DDF _k (p) is the deviation factor of point p, is the regularized standard deviation of point p. Both are expressed as:

Respectively expressed as:

Among them, N _kd (q) represents the k points closest to point q, and N _kd (p) represents the k points closest to point p. That is, through the relationship between point p and surrounding adjacent points, the probability that point p belongs to the abnormal point is deduced. And this probability value is converted into the distance calculation weight of point p and brought into the final objective function. Using the SVD algorithm, the objective function with weights can be minimized, and the rotation transformation R and translation transformation T between the two cross-sectional point clouds to be registered can be calculated.

The weight is introduced into the objective function as a constraint, and the weight is used to control the search direction of the optimization algorithm, so that the registration accuracy of the normal area can be effectively guaranteed, and the influence of abnormal points on the registration accuracy can be reduced. When scanning the subway tunnel in different periods, the position of the 3D laser scanner cannot be fixed, and the point cloud data sets of the two sections at the same location in the subway tunnel are in two different coordinate systems. Since the section of the subway tunnel is basically close to a circle, the center of the section point cloud can be preliminarily determined by least squares fitting of the ring. Initial registration of the cloud. Using the improved ICP algorithm, by introducing the concept of weight constraints and assigning higher weights to the more important regions in the registration point cloud, high-precision registration of the two-phase section point cloud can be obtained. Finally, the overall deformation of the tunnel section can be understood through the distance between the corresponding point pairs of the two point cloud datasets (see Figure 8 and Figure 9).

The embodiment of the present invention also provides a tunnel section deformation analysis device, the device is used to implement the method shown in the above embodiment, as shown in Figure 10, the device includes:

The collection unit 901 is used to collect tunnel section information to obtain point cloud data sets;

The comparative analysis unit 902 is configured to analyze and compare the point cloud data sets collected at different periods of the tunnel section to be detected, so as to determine the deformation trend of the tunnel section to be detected.

The section information of the tunnel in different periods can be visualized and displayed, and the deformation of the section can be determined through the comparison of the point cloud data sets, so as to realize the overall and comprehensive detection of the trend of the tunnel section over time. Moreover, the measurement accuracy of the present application is relatively high, and the measurement method is relatively simple and easy to implement.

In a possible implementation, the acquisition unit 901 is also used to collect tunnel section information by scanning the same tunnel section to be detected in different periods with a three-dimensional laser scanner; acquire the points of two periods according to the tunnel section information Cloud datasets P ₁ and P ₂ .

In a possible implementation, the comparative analysis unit 902 is also used to perform circle fitting on the point cloud datasets of the two periods in their respective coordinate systems, so as to determine the centers of the two point cloud datasets respectively; Place the two point cloud datasets in the same coordinate system, and then coincide the center of the circle to perform the first registration of the two point cloud datasets; perform the first registration on the two point cloud datasets after the first registration according to the ICP algorithm. The second registration: randomly select one of the point cloud datasets as the benchmark point cloud dataset, and use the ANN algorithm to determine the corresponding points of each benchmark point in the other point cloud dataset; analyze the relationship between each pair of benchmark points and The relationship between the corresponding points to determine the deformation trend of the tunnel section to be detected; where the reference point is the point in the reference point cloud data set.

It should be noted that, in this document, the term "comprising", "comprising" or any other variation thereof is intended to cover a non-exclusive inclusion such that a process, method, article or apparatus comprising a set of elements includes not only those elements, It also includes other elements not expressly listed, or elements inherent in the process, method, article, or device. Without further limitations, an element defined by the phrase "comprising a..." does not preclude the presence of additional identical elements in the process, method, article, or apparatus comprising that element.

The serial numbers of the above embodiments of the present invention are for description only, and do not represent the advantages and disadvantages of the embodiments.

Through the description of the above embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus a necessary general-purpose hardware platform, and of course also by hardware, but in many cases the former is better implementation. Based on such an understanding, the essence of the technical solution of the present invention or the part that contributes to the prior art can be embodied in the form of software products, and the computer software products are stored in a storage medium (such as ROM/RAM, disk, CD) contains several instructions to make a mobile terminal (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) execute the methods described in various embodiments of the present invention.

Embodiments of the present invention have been described above in conjunction with the drawings, but the present invention is not limited to the above-mentioned specific implementations, which are only illustrative and not restrictive. Under the enlightenment of the invention, many forms can also be made without departing from the gist of the present invention and the scope of protection of the claims, and these all belong to the protection of the present invention.

Claims (14)

1. a kind of tunnel cross-section deformation analytical method, which is characterized in that the method includes：

Step 1, tunnel cross-section information is acquired, to obtain point cloud data collection；

Step 2, the point cloud data collection that tunnel cross-section to be detected is acquired in different times is analyzed and is compared, described in determination The deformation tendency of tunnel cross-section to be detected.

2. according to the method described in claim 1, it is characterized in that, step 1 includes：

Step 11, it in such a way that three-dimensional laser scanner is scanned the same tunnel cross-section to be detected of different times, adopts Collect tunnel cross-section information；According to the point cloud data collection P in two periods of the tunnel cross-section acquisition of information₁And P₂。

3. according to the method described in claim 1, it is characterized in that, step 2 includes：

Step 21, round fitting is carried out respectively in respective coordinate system to the point cloud data collection in two periods, to determine two respectively The center of circle of a point cloud data collection；

Step 22, two point cloud data collection are placed under the same coordinate system, then overlap the center of circle, with to two point cloud data collection Carry out the first registration；

Step 23, two point cloud data collection after being registrated according to iteration closest approach ICP algorithm pair first carry out the second registration；

Step 24, a selected wherein phase point cloud data collection utilizes approximate KNN algorithm ANN as datum mark cloud data set at random It is concentrated in another phase point cloud data, determines the corresponding points of each datum mark respectively；

Step 25, the relationship for analyzing each pair of datum mark and corresponding points, with the deformation tendency of the determination tunnel cross-section to be detected；

Wherein, the datum mark is the point in the datum mark cloud data set.

4. according to the method described in claim 3, it is characterized in that,

The step 21, specifically includes：According to the point cloud data collection in two periods of least square method pair in respective coordinate system Round fitting is carried out respectively.

5. according to the method described in claim 3, it is characterized in that, step 23 includes：

Step 231, point cloud data collection P is calculated₂In each point in point cloud data collection P₁In correspondence neighbor point, to obtain One group of N number of corresponding points pair；

Step 232, it determines so that the distance of first group of N number of corresponding points pair and minimum translation parametersAnd rotation parameterIts In, the translation parametersWith the rotation parameterFor the parameter in rigid body translation；

Step 233, when the distance of first group of N number of corresponding points pair and when more than or equal to pre-determined distance, according to the point cloud data Collect P₂, the translation parametersWith the rotation parameterDetermine point cloud data collection P '₂；Calculate point cloud data collection P '₂In it is each A point is in point cloud data collection P₁In correspondence neighbor point, to obtain second group of N number of corresponding points pair；

Step 234, it when the distance of second group of N number of corresponding points pair and when more than or equal to the pre-determined distance, redefines new Point cloud data collection is simultaneously iterated calculating, until the distance of i-th group of corresponding points pair is less than pre-determined distance.

6. according to the method described in claim 5, it is characterized in that,

The step 231, specifically includes：According to preset condition, and using KD-tree structures in P₁Middle determining P₂Each point Correspondence point of proximity, composition corresponding points to set { (P_{1, i}, P_{2, i}| i=1,2 ..., N) }；

Wherein, preset condition is：ε(P₁, P₂)min(d²(P_{1, i}P_2,i))；P_{1, i}With P_{2, i}For i-th group of corresponding points pair, P_{1, i}Belong to P₁, P_{2, i}Belong to P₂。

7. according to the method described in claim 5, it is characterized in that,

The step 232, specifically includes：Step 2321, by N number of corresponding points to substituting into object function；It is true by iterative algorithm Determine rigid body translation matrixTo determine so that the distance of corresponding points pair and minimum translation parametersAnd rotation parameter

Wherein, target function value be N number of corresponding points pair distance and, formula is：

Wherein, N P₁The number at midpoint, P_{1, i}With P_{2, i}For i-th group of corresponding points pair, P_{1, i}Belong to P₁, P_{2, i}Belong to P₂。It is 3 × 1 Translation matrix,For 3 × 3 spin matrix.

8. according to the method described in claim 5, it is characterized in that,

Target function value be N number of corresponding points pair distance and, formula is：

Wherein, N P₁The number at midpoint, P_{1, i}With P_{2, i}For i-th group of corresponding points pair, P_{1, i}Belong to P₁, P_{2, i}Belong to P₂。It is 3 × 1 Translation matrix,For 3 × 3 spin matrix, w_iTo constrain weight factor.

9. according to the method described in claim 8, it is characterized in that,

Wherein, DDF_k(p) deviation factors for being point p,For the regularization standard deviation of point p.

10. according to the method described in claim 5, it is characterized in that,

The step 231 further includes：Delete the corresponding points pair that distance is more than distance threshold.

11. according to the method described in claim 5, it is characterized in that,

The distance of corresponding points pair is projector distance of the Euclidean distance of point-to-point transmission on the radius extended line for overlapping the center of circle.

12. a kind of tunnel cross-section deformation analysis device, which is characterized in that described device is appointed in requiring 1 to 11 for perform claim Method described in meaning one, described device include：

Collecting unit, for acquiring tunnel cross-section information, to obtain point cloud data collection；

Comparative analysis unit, the point cloud data collection for acquiring tunnel cross-section to be detected in different times carry out analysis and it is right Than with the deformation tendency of the determination tunnel cross-section to be detected.

13. device according to claim 12, which is characterized in that

The collecting unit is additionally operable to sweep the same tunnel cross-section to be detected of different times by three-dimensional laser scanner The mode retouched acquires tunnel cross-section information；According to the point cloud data collection P in two periods of the tunnel cross-section acquisition of information₁And P₂。

14. device according to claim 12, which is characterized in that

The comparative analysis unit is additionally operable to carry out the point cloud data collection in two periods respectively in respective coordinate system round quasi- It closes, to determine the center of circle of two point cloud data collection respectively；Two point cloud data collection are placed under the same coordinate system, then by the center of circle It overlaps, to carry out the first registration to two point cloud data collection；Two points after being registrated according to iteration closest approach ICP algorithm pair first Cloud data set carries out the second registration；A selected wherein phase point cloud data collection is as datum mark cloud data set at random, most using approximation Nearest neighbor algorithm ANN is concentrated in another phase point cloud data, determines the corresponding points of each datum mark respectively；Analyze each pair of datum mark with it is right The relationship that should be put, with the deformation tendency of determination tunnel cross-section to be detected；Wherein, the datum mark is the datum mark cloud data set In point.