High-rise building deformation displacement extraction method based on improved CEEMD endpoint effect

Abstract

The invention discloses a CEEMD high-rise building deformation displacement extraction method based on an improved endpoint effect, and the method comprises the following steps: obtaining dynamic positioning coordinate data through a GNSS satellite receiver disposed at a building monitoring point; and noise removal and building displacement information decomposition and extraction are carried out by utilizing 4G transmission and using a CEEMD algorithm with an improved end effect, and wind swing, inclination, settlement and horizontal displacement in building deformation displacement are calculated and recorded. The method for extracting the deformation displacement of the high-rise building is good in real-time performance and high in accuracy, the existing high-rise building can be conveniently monitored for a long time, the tedious problem that manual measurement is still needed for extracting the deformation information of the building at the present stage is solved, and a large amount of manpower and material resources can be saved.

Description

Technical Field

[0001] The invention relates to the field of building safety monitoring, and in particular to a method for suppressing EMD endpoint effect. Technical Background

[0002] Building safety monitoring, that is, monitoring the deformation caused by changes in its hydrogeological conditions, soil properties, climate environment, building load, structural characteristics, etc., is divided into external deformation monitoring and internal deformation monitoring. External deformation monitoring mainly measures the changes in the three-dimensional geometric shape of the deformed body in space; internal deformation monitoring mainly uses various special sensors to observe the strain, stress, temperature, seepage pressure, soil pressure, pore pressure, and expansion joint opening and closing of the deformed body structure. Since internal deformation monitoring can only provide local deformation information and usually requires pre-embedded sensors during the construction period, most completed buildings such as houses do not have observation conditions. Therefore, external deformation monitoring has become an irreplaceable main monitoring method.

[0003] Looking at the development of deformation monitoring technology at home and abroad over the past decades, the main methods for monitoring the external deformation of buildings include conventional ground measurement methods using total stations and levels, photogrammetry, GPS measurement, etc. At present, the main means of monitoring is still manual measurement, that is, surveyors set up various measuring instruments on site, manually observe and collect data, and then calculate deformation information through industry processing. With the increasing number of monitored buildings and the gradual improvement of monitoring requirements, the disadvantages of manual monitoring methods, such as low work efficiency, slow information feedback, and inconvenience for cluster management, are becoming increasingly apparent.

[0004] EMD (Empirical Mode Decomposition) is an adaptive signal processing method proposed by Huang et al. of NASA in 1998. The IMF components obtained by decomposition have real physical meanings, so they are widely used to process nonlinear and non-stationary signals. The traditional EMD method is prone to the problem of endpoint effect in the data extraction process. In order to improve this problem, the academic community has proposed a variety of algorithms such as the mirror continuation method, but such methods are mostly used for signals with obvious periodicity and are not completely suitable for the processing of building deformation and displacement data. Therefore, according to the characteristics of building deformation and displacement data, the central symmetric continuation method is proposed to curb the endpoint effect during modal decomposition. Summary of the invention

[0005] Purpose of the invention: The purpose of the present invention is to address the deficiencies in the prior art and propose a method for extracting deformation and displacement of high-rise buildings based on the improved CEEMD endpoint effect, which can obtain the deformation and displacement of buildings in real time to make up for the deficiencies in the current housing safety status monitoring methods.

[0006] Technical solution: According to a first aspect of the present invention, a method for suppressing EMD endpoint effect is provided, comprising the following steps:

[0007] like Figure 4 As shown in , find a set of maximum points Q(t l , x max (t l )) and the minimum point E(t2, x min (t2)); This set of extreme points is divided into endpoints P(t p , x p ) is the symmetric center and extends to the outside of the time series. From the maximum value Q(t l , x max (t l )) The reference point after the central symmetric extension is the new minimum point Q′(2·t p -t l , 2·t p -x max (t l )), by the minimum value E(t2,x min (t2)) The reference point after the central symmetric extension is the new maximum point E′(2·t p -t2,2·t p -x min (t2)); the position of the new extreme point is the new reference point of the original signal when performing CEEMD. The new extreme point after extension and the extreme point of the original signal are used as interpolation points to perform cubic spline function fitting to improve the problem of envelope divergence at the endpoints.

[0008] According to a second aspect of the present invention, there is provided a calculation method for determining horizontal movement, tilt, and settlement of a building using GNSS receiver coordinate signals, comprising the following steps:

[0009] Assuming that three GNSS satellite receiver monitoring points A, B, and C are installed on the roof of the same building, and the three receivers are not collinear, the original coordinates are extracted by CEEMD to remove noise and wind swing information to obtain a more stable displacement. The more stable displacement of the i-th dimension data received by the K-th monitoring point in the same time period ΔT is recorded as i k (t), we can get x A (t), x B (t), x C (t), y A (t), y B (t), y C (t), z A (t), z B (t), z C (t) A total of 9 stable displacement signals;

[0010] (1) Tilt angle: The tilt angle is calculated by the normal change of the plane composed of monitoring points. The normal vector l determined at t = 0 l The expression (x, y, z) satisfies

[0011] (x A (0)-x B (0))+(y A (0)-y B (0))+(z A (0)-z B (0))=0

[0012] (x A (0)-x C (0))+(y A (0)-y C (0))+(z A (0)-z C (0))=0

[0013] The normal vector l2(x′, y′, z′) determined at time t=ΔT satisfies the expression

[0014] (x A (ΔT)-x B (ΔT))x′+(y A (ΔT)-y B (ΔT))y′+(z A (ΔT)-z B (ΔT))z′=0

[0015] (x A (ΔT)-x C (ΔT))x′+(y A (ΔT)-y C (ΔT))y′+(z A (ΔT)-z C (ΔT))z′=0

[0016] Then the inclination angle θ = cos -1 ((l1·l2) / (|l1|×|l2|)).

[0017] (2) Settlement: Settlement refers to the change in the measured height z of a building, which includes changes in the building height caused by thermal expansion and contraction, as well as building sinking caused by insufficient foundation strength. If irregular settlement occurs, it will also be reflected in the tilt deformation. Here, the average value of the z changes at the three monitoring points is used as the general settlement.

[0018] Δh=(z A (ΔT)-z A(t)+z B (ΔT)-z B (t)+z C (ΔT)-z C (t)) / 3

[0019] (3) Horizontal displacement: Horizontal displacement refers to the change in the position of the building on the xy plane. Here, the average value of the x and y changes of the three monitoring points is used as the horizontal displacement value.

[0020] Δx=(x A (ΔT)-x A (t)+x B (ΔT)-x B (t)+x C (ΔT)-x C (t)) / 3

[0021] Δy=(y A (ΔT)-y A (t)+y B (ΔT)-y B (t)+y C (ΔT)-y C (t)) / 3

[0022] Δd=(Δx,Δy)

[0023] Beneficial effects: Compared with the prior art, the endpoint effect suppression method used in the present invention is designed based on the characteristics of the signal itself, has strong practicability and high accuracy, can record the dynamic positioning information of building monitoring points around the clock, has good real-time performance, and is convenient for long-term monitoring of existing high-rise buildings. It solves the cumbersome problem of manual measurement required to extract building deformation information at this stage, and can save a lot of manpower and material resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 Schematic diagram of the distribution of monitoring points on the roof of a high-rise building in the present invention.

[0026] Figure 2 This is a schematic diagram of the distribution of reference base stations that are arranged in coordination with monitoring points in the present invention.

[0027] Figure 3 This is a flow chart of the implementation of the high-rise building deformation and displacement extraction algorithm of the present invention.

[0028] Figure 4 It is a schematic diagram of the centrosymmetric extension method for suppressing the endpoint effect in the present invention.

[0029] Figure 5 This is the extraction effect of the present invention on the wind sway displacement of a high-rise building. Specific implementation plan

[0030] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0031] Step 1: Arrange monitoring points and base stations on the building. In the embodiment, the test building is measured by the total station to be 41m long, 36m wide, and 103m high. The distribution diagram of the three monitoring points on the roof is as follows: Figure 1 As shown in the figure, the satellite positioning receiver used at the monitoring point is BD-100. The BD-100 Beidou locator is a global star navigation system locator that supports the Beidou navigation system and the GPS navigation system. It has a full range of real-time positioning functions and can meet professional positioning requirements. The three ground base stations are distributed as follows: Figure 2 As shown, the equipment used in the base station is Huace H3.

[0032] Step 2: After setting up the monitoring points and base stations, use the static relative positioning method to obtain the positioning coordinates of the monitoring points in real time. The data frequency is divided into 1 / second and 1 / hour, which are used for wind pendulum displacement extraction and tilt, settlement, and horizontal displacement extraction respectively; the data is sent to the cloud database via a 4G card for storage, and the CEEMD algorithm with improved endpoint effect is used to extract wind pendulum displacement from positioning data with a receiving frequency of 1 / second and to extract tilt, settlement, and horizontal displacement from positioning data with a receiving frequency of 1 / hour. The positioning data with a frequency of 1 / hour received at the three monitoring points for 26 working days are intercepted for deformation displacement extraction, and the positioning data with a receiving frequency of 1 / second within 30 consecutive minutes are used for deformation displacement extraction. The x-coordinate component of one of the monitoring points extracts the wind pendulum displacement.

[0033] Step 3: Perform CEEMD decomposition on the acquired data, add 100 groups of white noise with opposite signs to the original signal to form 100 groups of signals to be decomposed, use EMD to decompose the 100 groups of signals into multi-level IMFs and take the average value of each level of IMF as the final IMF result. When performing EMD, the central symmetric extension method is used to obtain the interpolation points of the maximum and minimum envelopes to suppress the endpoint effect.

[0034] Step 4: Record and analyze the start and end coordinate positions of the trend items remaining after all IMFs are decomposed within the 26-day period. Calculate the average change of the three monitoring points and the change of the normal angle of the plane where they are located through the known position relationship of the monitoring points. The specific values ​​are shown in Table 1. Record and analyze the total IMF information remaining after removing the first three IMFs within a continuous 30-minute period under the working state. Figure 5 The figure shows the extraction effect of x-dimensional positioning information of monitoring point A during this time period.

[0035] Table 1 Deformation and displacement of target buildings from November 15 to December 10, 2022

[0036]

Claims

1. A method for extracting deformation and displacement of high-rise buildings based on improved CEEMD endpoint effect, characterized in that: The CEEMD modal decomposition algorithm is used to decompose the GNSS data of the coordinate monitoring points of high-rise buildings to remove noise, extract wind sway displacement, and three deformation data of building settlement, tilt, and horizontal displacement, and finally display the building safety status on the visualization platform; the analysis method includes the following steps: Step 1: Set up a GNSS satellite positioning receiver on the roof of a high-rise building, and use a differential positioning algorithm with the base station to obtain the three-dimensional positioning coordinates of the fixed point on the roof of the building. The data receiving frequency mode of each receiver is compatible with 1 / second and 1 / hour; Step 2: Perform CEEMD empirical mode decomposition on each dimensional coordinate time series signal sequence of each receiver, and use the central symmetric extension method to suppress the endpoint effect in each signal decomposition to obtain the IMF signal component; Step 3: Classify and merge the IMF signal components at all levels according to the change frequency and total displacement direction of the IMF signal components of each dimensional coordinate signal to determine whether they belong to wind sway displacement, and determine whether the building has obvious settlement, tilt, and horizontal displacement deformation data based on the trend item.

2. The high-rise building deformation displacement extraction algorithm based on improved CEEMD endpoint effect according to claim 1 is characterized in that: The GNSS technology used for building deformation monitoring adopts static relative positioning to weaken the impact of electromagnetic waves when passing through the ionosphere. It is necessary to deploy reference base stations (referred to as base stations) to form a relative baseline with the monitoring points. The monitoring points are deployed on the main structure of the building according to the needs of the monitoring points. Generally, it is advisable to deploy 3 monitoring points on the top of each building, and each monitoring point is equipped with a set of GNSS receiving equipment (sensor). The base station provides the monitoring coordinate calculation benchmark for the monitoring point. In order to improve the accuracy of the differential and reduce the impact of abnormal changes in GPS elevation, the base station should be deployed to ensure that: (1) The base station is located in an open and unobstructed location. The elevation angle of obstacles in the field of view should not be greater than 15°. There should be no radio interference nearby to avoid GPS signal loss. (2) The location of the base station is stable to avoid location changes caused by settlement or other ground deformation; (3) The base station should be at least 1 meter above the ground to suppress the multipath effect.

3. The high-rise building deformation displacement extraction algorithm based on improved CEEMD endpoint effect according to claim 1 and claim 2 is characterized in that: The wind-sway displacement in the deformation displacement of high-rise buildings is the main signal, which does not have an obvious periodic swing law in its change. Therefore, without prior conditions, it is determined that the trend term of the signal at the endpoint is linear. Before using CEEMD, a group of maximum points (t1, x max (t1))) and the minimum point (t2, x min (t2)), this set of extreme points is divided into endpoints P(t P , x P ) is the symmetric center and extends to the outside of the time series, from the maximum value (t1, x max (t1))) The reference point after the central symmetric extension is the new minimum point (2·t P -t1,2·x P -x max (t1)), by the minimum value (t2, x min (t2)) The reference point after the central symmetric extension is the new maximum point (2·t P -t2,2·x P -x min (t2))), its position is the new reference point of the original signal when performing CEEMD. The new extreme point after extension and the extreme point of the original signal are used as interpolation points to perform cubic spline function fitting to improve the endpoint effect.

4. The high-rise building deformation displacement extraction algorithm based on improved CEEMD endpoint effect according to claim 3 is characterized in that: (1) Add K groups of positive and negative paired auxiliary white noises to each dimensional component of the positioning coordinate time series signal x(t), that is, the signals obtained after adding positive and negative auxiliary white noises are y1(t) and y2(t), y1(t) = x(t) + n(t), y2(t) = x(t) - n(t); where x(t) is the original signal, n(t) is the auxiliary Gaussian white noise with a mean of zero, and the number of signals obtained is 2K; (2) For the 2K signals obtained in (1), all extreme points are obtained respectively, and the new extreme reference points of the signals are obtained by the central symmetric extension method. The envelopes of the maximum and minimum points are constructed by the cubic spline interpolation curve and their mean values ​​c(t) = [max(t) + min(t)] / 2. An IMF component can be obtained, where max(t) is the upper envelope of the signal x(t) and min(t) is the lower envelope of the signal x(t). The jth component obtained by decomposing the i-th signal is denoted as c ij ; (3) Sum and average the 2K groups of corresponding IMF components obtained in (2) to obtain the IMF components c of each order after the EMD decomposition of the signal x(t). j , (4) Analyze the IMFs of each level obtained in (3). If the input data is collected at a frequency of 1 / second, the first, second, and third level IMFs are removed as noise, and the trend term is removed as deformation displacement. The remaining signal is used as wind pendulum displacement. If the input data is collected at a frequency of 1 / hour, the last level IMF is used as the displacement change occurring at the monitoring point to calculate tilt, settlement, and horizontal displacement.

Images