CN-121977609-A - Static leveling correction system and method based on vibration compensation
Abstract
The invention provides a static leveling correction system and a static leveling correction method based on vibration compensation, wherein a data processing system and a static leveling instrument arranged on a measuring point of a building structure are adopted, and vertical acceleration data, leveling data and temperature data are used as a data sample; and the data processing system respectively adopts different relation models to estimate and obtain different leveling errors according to the vertical acceleration data and the temperature data acquired in real time, and dynamically adjusts the relation model or directly obtains the compensated leveling value according to the deviation condition of the different leveling errors. The invention can remarkably improve the static leveling accuracy and reliability in a vibration environment, can monitor and correct the measurement error in real time and improves the working efficiency.
Inventors
- ZHAO YIMING
- PAN XI
- ZHANG LONGLONG
- SONG JIALING
- CHEN HAO
Assignees
- 上海建工集团股份有限公司
Dates
- Publication Date
- 20260505
- Application Date
- 20251231
Claims (10)
- 1. A static leveling correction system based on vibration compensation is characterized by comprising a data processing system and a static leveling instrument arranged on a measuring point of a building structure, wherein the static leveling instrument is internally provided with a vertical acceleration sensor and a temperature sensor besides a pressure sensing function of measuring the liquid level height, and the static leveling instrument, the vertical acceleration sensor, the static leveling instrument and the temperature sensor are arranged on the building structure, The temperature sensor is respectively connected with and controlled by the data processing system in a communication way, the data processing system is deployed on the central server and used for data acquisition and error correction, the central time t c of the acquisition unit time period (t 1 ~t 2 ) of the vertical acceleration sensor is used as the acquisition time of the liquid level height of the static level and the acquisition time of the temperature, and the vertical acceleration data of the unit time period (t 1 ~t 2 ), The leveling data at the center moment t c and the temperature data at the center moment t c are taken as a data sample for carrying out subsequent leveling error correction, a relation model M 1 among vertical acceleration data, temperature data and leveling error and a relation model M 2 among independent acceleration data and leveling error are established in a data processing system based on a large number of data samples, and the data processing system acquires the vertical acceleration data, the leveling error and the leveling error in real time, The data processing system is used for processing the vertical acceleration data according to the vertical acceleration data acquired in real time, And the leveling data and the temperature data are respectively estimated by adopting different relation models M 1 and M 2 to obtain leveling errors epsilon 1 and epsilon 2 , and the relation models M 1 and M 2 are dynamically adjusted or the compensated leveling values are directly obtained according to the deviation conditions of the leveling errors epsilon 1 and epsilon 2 obtained by estimating the different relation models M 1 and M 2 .
- 2. The vibration compensation-based hydrostatic leveling correction system as set forth in claim 1, wherein the liquid level of the hydrostatic level is measured by a pressure sensor, the vertical acceleration is measured by a vertical fixed acceleration sensor, the temperature is measured by a resistance temperature sensor, and the pressure sensor, the acceleration sensor and the resistance temperature sensor are all monitored at the same position to ensure the accuracy of leveling correction.
- 3. The vibration-compensation-based static leveling correction system as set forth in claim 1 wherein prior to establishing the relational models M 1 and M 2 , the vertical acceleration data vector X 1 in the data samples is manually pre-processed, including denoising and filtering, to obtain a clear vibration signal vector X 2 , converting the vibration signal vector X 2 from the time domain to the frequency domain by fourier transform, and analyzing the fundamental frequency and corresponding amplitude of the vibration signal vector X 2 .
- 4. The vibration-compensation-based static leveling correction system as set forth in claim 1 wherein prior to establishing the relational models M 1 and M 2 , the level data Y 1 in the data sample is manually preprocessed, including extremum removal, denoising and smoothing, to obtain clear level data Y 2 , and the difference is taken as the leveling error epsilon by subtracting the preprocessed level data Y 2 from the level data Y 1 in the data sample.
- 5. The vibration-compensation-based static leveling correction system as set forth in claim 4 wherein a linear regression is employed to establish a linear regression equation epsilon '=a x t+b between the temperature measurement value T and the leveling error epsilon in the data sample, where a, b are regression coefficients and epsilon' is an estimate of the leveling error based on the temperature T and the linear regression equation, prior to establishing the relationship models M 1 and M 2 .
- 6. The vibration-compensation-based static leveling correction system as set forth in claim 4, wherein the method for establishing the relationship model M 1 among the vertical acceleration data, the temperature data and the leveling error is that the input parameters are { maximum value, minimum value, mean value, variance, base frequency, corresponding amplitude } of the vibration signal vector X 2 and the temperature data T, the output data is the leveling error epsilon, the relationship model M 1 is obtained by training the input parameters and the output data calculated based on a plurality of data samples by adopting a random forest method, and the relationship model M 1 represents the comprehensive influence degree of vibration and temperature on the leveling error.
- 7. The vibration-compensation-based static leveling correction system as set forth in claim 5, wherein the method for establishing the relationship model M 2 between the vertical acceleration data and the leveling error is that the input parameters are { maximum value, minimum value, mean value, variance, basic frequency, corresponding amplitude of the vibration signal vector X 2 }, the output data are leveling errors caused by vibration, the leveling error caused by vibration is a difference value of leveling error epsilon minus a leveling error estimated value epsilon' based on a temperature T and a linear regression equation, the relationship model M 2 is obtained by training based on the input parameters and the output data calculated by a plurality of data samples by a random forest method, and the relationship model M 2 represents the influence degree of individual vibration on the leveling error after the temperature linear compensation.
- 8. The vibration compensation-based static leveling correction system according to claim 1, wherein leveling errors epsilon 1 and epsilon 2 are estimated by using different relation models M 1 and M 2 respectively according to vertical acceleration data, temperature data and leveling data acquired in real time, wherein epsilon 1 is a comprehensive leveling error caused by vibration and temperature, epsilon 2 is a leveling error caused by vibration and is overlapped with an error value of temperature linear compensation estimation, a deviation threshold value of leveling errors calculated by different models is set as delta, the deviation threshold value is set according to 2 times of required leveling accuracy, the problem of poor leveling accuracy of the estimation is considered for the case of epsilon 1 -ε 2 delta, average values of epsilon 1 and epsilon 2 are used as leveling errors epsilon of a current data sample, the relation models M 1 and M 2 are dynamically updated by using a random forest method according to the calculation method, and the average values of epsilon 1 and epsilon 2 are used as final calculated leveling errors and are overlapped with actual leveling measurement values after the actual leveling compensation measurement values are obtained.
- 9. The static leveling correction method based on vibration compensation is characterized by comprising the following steps of: The static leveling correction system based on vibration compensation is installed and comprises a data processing system and a static leveling instrument arranged on a measuring point of a building structure, wherein the static leveling instrument is provided with a vertical acceleration sensor and a temperature sensor besides a pressure sensing function of measuring the liquid level height; the static leveling instrument, the vertical acceleration sensor and the temperature sensor are respectively in communication connection with and controlled by a data processing system, and the data processing system is deployed on a central server and is used for data acquisition and error correction; Taking the central time t c of the acquisition unit time period (t 1 ~t 2 ) of the vertical acceleration sensor as the acquisition time of the liquid level height and the acquisition time of the temperature of the static level, and taking the vertical acceleration data of the unit time period (t 1 ~t 2 ), the level data of the central time t c and the temperature data of the central time t c as a data sample for carrying out subsequent leveling error correction; Based on a large number of data samples, a relation model M 1 among vertical acceleration data, temperature data and leveling errors and a relation model M 2 among independent acceleration data and leveling errors are established in a data processing system; The data processing system acquires vertical acceleration data, leveling data and temperature data in real time, estimates leveling errors epsilon 1 and epsilon 2 by adopting different relation models M 1 and M 2 respectively according to the vertical acceleration data, the leveling data and the temperature data acquired in real time, and dynamically adjusts the relation models M 1 and M 2 or directly obtains compensated leveling values according to deviation conditions of leveling errors epsilon 1 and epsilon 2 estimated by the different relation models M 1 and M 2 .
- 10. The vibration compensation-based static leveling correction method as set forth in claim 9, wherein leveling errors epsilon 1 and epsilon 2 are estimated by using different relation models M 1 and M 2 respectively according to vertical acceleration data, temperature data and leveling data acquired in real time, wherein epsilon 1 is a comprehensive leveling error caused by vibration and temperature, epsilon 2 is a leveling error caused by vibration and is overlapped with an error value of temperature linear compensation estimation, a deviation threshold value of leveling errors calculated by different models is set as delta, the deviation threshold value is set according to 2 times of required leveling accuracy, the problem of poor leveling accuracy of the estimation is determined according to the condition of epsilon 1 -ε 2 delta, the average value of epsilon 1 and epsilon 2 is used as leveling error epsilon of a current data sample, the relation models M 1 and M 2 are dynamically updated by using a random forest method according to the calculation method, and the average value of epsilon 1 and epsilon 2 is used as a final calculated leveling error and is overlapped with an actual leveling measurement value after leveling compensation is carried out.
Description
Static leveling correction system and method based on vibration compensation Technical Field The invention belongs to the field of civil engineering building construction informatization, and particularly relates to a static leveling correction system and method based on vibration compensation. Background Engineering structure deflection monitoring (such as vertical deformation of a bridge girder body under a vehicle load) or vertical relative displacement monitoring (such as building settlement) generally adopts a static level instrument. The working principle of the device is based on the principle of a communicating vessel, wherein a pipe connecting a plurality of measuring points is filled with liquid, and when all measuring points are stationary, the liquid level of each point is the same. If a certain measuring point is displaced vertically (up or down), the liquid level at that point will change accordingly. By detecting this change in level (typically by measuring the column pressure with a pressure sensor), the amount of deformation or settlement of the point relative to the datum point can be calculated. However, in an environment with vibration, the liquid in the communicating pipe is easy to be impacted, so that the liquid level generates waves or shakes (such as the liquid oscillates back and forth). The shaking can interfere with the stable detection of the liquid level height, so that measurement deviation, which is usually represented as noise (small-amplitude random fluctuation) in data, can even appear obvious data jump or large fluctuation when the vibration is strong, and seriously affects the accuracy and precision of the monitored data. Meanwhile, in an environment with temperature change, it is generally considered that there is a linear relationship between the temperature and the liquid level of the static level (i.e. when the temperature increases or decreases, the liquid level will increase or decrease uniformly due to expansion or contraction). This simple linear relationship is broken when vibrations are present in the environment at the same time. The reason is that the actual change in the level is now a result of the superposition of temperature effects and vibration effects-the temperature causes a slow, steady rise and fall of the liquid level, whereas the vibration causes a rapid sloshing and momentary fluctuation of the liquid level. These two effects are mixed together, and it is difficult to distinguish them separately, so that the liquid level reading cannot be linearly corrected by simply using temperature, and the measurement accuracy is significantly affected. These problems limit the reliable use of static levels in dynamic environments. Disclosure of Invention The invention aims to provide a static leveling correction system and a static leveling correction method based on vibration compensation, which improve the static leveling precision through the compensation of a vibration signal on a static leveling error. In order to solve the technical problems, the invention provides the following technical scheme: A static leveling correction system based on vibration compensation comprises a data processing system and a static leveling instrument arranged on a measuring point of a building structure, wherein the static leveling instrument is internally provided with a vertical acceleration sensor and a temperature sensor besides a pressure sensing function of measuring the liquid level height, and the static leveling instrument, the vertical acceleration sensor, The temperature sensor is respectively connected with and controlled by the data processing system in a communication way, the data processing system is deployed on the central server and used for data acquisition and error correction, the central time t c of the acquisition unit time period (t 1~t2) of the vertical acceleration sensor is used as the acquisition time of the liquid level height of the static level and the acquisition time of the temperature, and the vertical acceleration data of the unit time period (t 1~t2), The leveling data at the center moment t c and the temperature data at the center moment t c are taken as a data sample for carrying out subsequent leveling error correction, a relation model M 1 among vertical acceleration data, temperature data and leveling error and a relation model M 2 among independent acceleration data and leveling error are established in a data processing system based on a large number of data samples, and the data processing system acquires the vertical acceleration data, the leveling error and the leveling error in real time, The data processing system is used for processing the vertical acceleration data according to the vertical acceleration data acquired in real time, And the leveling data and the temperature data are respectively estimated by adopting different relation models M 1 and M 2 to obtain leveling errors epsilon 1 and epsilon 2, and the relat