Automatic observation method and platform for dam vertical displacement monitoring reference network based on GNSS

Abstract

The application discloses a kind of based on GNSS's dam vertical displacement monitoring reference network automation observation method.It includes the following steps, step one: point site reconnaissance, instrument and observation pier selection;Step two: GNSS data preprocessing and scheme optimization;In GNSS vertical displacement automation monitoring, determine the optimal observation period, repair frequent small cycle jump;Step three: baseline solution and network adjustment method;In GNSS vertical displacement automation monitoring, troposphere delay correction, thermal expansion effect correction these models are applied to baseline solution, introduce prior height difference information in network adjustment, improve the accuracy and reliability of baseline solution.The application has the advantages of improving the vertical direction of GNSS baseline solution precision.The application also discloses a kind of based on GNSS's dam vertical displacement monitoring reference network automation observation platform.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of dam deformation monitoring automation, more particularly, it is a kind of GNSS-based dam vertical displacement monitoring reference network automation observation method.The present application also relates to the observation platform used by the GNSS-based dam vertical displacement monitoring reference network automation observation method. BACKGROUND

[0002] In dam safety monitoring, the vertical displacement reference network is very important, which provides a reference for dam vertical displacement monitoring and monitors the deformation of the downstream surface of the dam, etc.According to the specification, the observation of the vertical displacement reference network of the concrete dam or the earth-rock dam is generally measured according to the national first or second leveling measurement accuracy, which cannot realize automatic measurement and consumes a large amount of manpower and material resources.With the development of intelligent management of hydropower stations, the observation automation of the dam vertical displacement monitoring reference network has become a relatively prominent short board in the whole dam safety monitoring system.The traditional static leveling observation facilities can realize the automatic monitoring of vertical displacement, but the degree of automation is not high, the equipment is expensive, the applicable conditions are harsh, the maintenance cost is high, and the monitoring range is small, which is difficult to meet the demand of dam vertical displacement monitoring automation.

[0003] At present, with the formal completion of the third generation of Beidou system, it is possible to monitor the millimeter-level GNSS dam vertical displacement reference network based on domestic Beidou high-precision receivers.On the one hand, GNSS technology has the monitoring capabilities of all-weather, real-time, automation, and high sampling rate;on the other hand, the use of domestic Beidou GNSS receivers for dam vertical displacement reference network monitoring has low cost, simple installation and operation, high automation, and flexible monitoring point layout, which can significantly reduce the construction cost of dam vertical displacement automation monitoring.However, under the existing hardware conditions of domestic receivers and antennas, the use of universal GNSS data processing methods can only make the vertical direction accuracy of baseline vector solution reach 3-5mm, which is difficult to replace second or even first leveling measurement, and difficult to meet the demand of intelligent development of hydropower station management.

[0004] Therefore, it is necessary to develop a GNSS-based dam vertical displacement monitoring reference network automation observation method and platform that can improve the vertical direction solution accuracy of GNSS baseline and meet the demand of intelligent development of hydropower station management. SUMMARY

[0005] The first object of the present application is to provide a GNSS-based dam vertical displacement monitoring reference network automatic observation method, which optimizes the point arrangement scheme, adopts a precise dual-frequency Beidou receiver and antenna equipment, standardizes the observation pier embedding process, simultaneously adopts a GNSS precise data solving method and adjustment strategy based on the Beidou technology, and improves the vertical direction solving precision of the GNSS baseline from 3-5 mm to 1-2 mm, and even sub-millimeter level in the case that the observation condition is good (i.e. the observation quality is good, including good point selection, stable foundation, no obstruction above the station, high observation quality, good receiver antenna hardware, etc.), which can replace the first / second order leveling, and realize the automatic monitoring of the dam vertical displacement monitoring system, and meet the intelligent development needs of the hydropower station management.

[0006] The second object of the present application is to provide a GNSS vertical displacement monitoring automatic platform used in the GNSS-based dam vertical displacement monitoring reference network automatic observation method, which can realize the automatic monitoring of the dam vertical displacement under the premise of meeting the millimeter level monitoring precision, and the precision is expected to replace the current second order or even first order leveling.

[0007] In order to realize the first object of the present application, the technical scheme of the present application is as follows: a GNSS-based dam vertical displacement monitoring reference network automatic observation system, characterized by comprising the following steps,

[0008] Step one: point selection, instrument selection and observation pier selection;

[0009] Optimizing the GNSS network shape design and point selection, adopting high-precision Beidou receiver and antenna instrument equipment, and optimizing the observation pier foundation and structure design, create good hardware conditions and observation environment for GNSS vertical displacement automatic monitoring;

[0010] Step two: GNSS data preprocessing and scheme optimization;

[0011] In the GNSS vertical displacement automatic monitoring, a precise data preprocessing method is adopted to weaken the influence of multipath effect, retain high-quality GNSS observation values, determine the optimal observation period, and repair frequent small cycle slips;

[0012] Step three: baseline solving and network adjustment method;

[0013] In the GNSS vertical displacement automatic monitoring, the models such as troposphere delay correction and thermal expansion effect correction are applied to the baseline solving, and the prior height difference information is introduced in the network adjustment, which improves the precision and reliability of the baseline solution for calculating the dam displacement.

[0014] The application proposes a set of establishment methods of the dam vertical displacement monitoring reference network automation system based on GNSS technology from the aspects of monitoring point site selection, receiver antenna hardware selection, observation pier structure design, GNSS precise data processing model and method optimization, and realizes the automatic monitoring of the dam vertical displacement reference network based on GNSS technology.

[0015] In the above technical solution, in step one, point site selection, instrument selection and observation pier selection, specifically including the following steps:

[0016] Step 11: Optimize GNSS network design to improve point site selection standards;

[0017] In the selection of dam vertical displacement monitoring reference network points, in addition to following the point selection principles of the monitoring reference network in the current national and industry standards and specifications such as Technical Standard for Safety Monitoring of Concrete Dams (GB_T 51416-2020) and Technical Specification for Safety Monitoring of Earth and Rockfill Dams (SL 551-2012), the reference network points should be selected as far as possible from open water surface and high-voltage towers in higher places, while ensuring that there are no mountains, large trees and other significant obstructions within a 30-degree height angle in the zenith direction;

[0018] Step 12: Use high-precision dual-frequency Beidou receivers and choke coil antenna equipment;

[0019] Step 13: Ensure the stability of the observation pier foundation and optimize the structure design of the observation pier and its associated facilities;

[0020] The observation pier should be set on bare bedrock, and the observation pier foundation should be anchored in the bedrock more than 1 meter below with sufficient amount of steel bars; if the bedrock is located at a deep position below the ground, a steel pipe pile should be used to reach at least 5 meters below the original soil, and concrete should be filled for fixation.

[0021] In the above technical solution, in step two, GNSS data preprocessing and scheme optimization, the specific method is:

[0022] Step 21: Weaken the influence of multipath effect;

[0023] The choke coil antenna can absorb more than 90% of the satellite signals reflected by the ground with its diameter suppression plate, significantly weakening the error caused by the multipath effect, and improving the baseline solution quality and precision;

[0024] The solar panels and other associated facilities should not be higher than the bottom of the GNSS choke coil antenna to further weaken the influence of the multipath effect;

[0025] Step 22: Remove low-quality satellite observation values, and select the optimal observation period for solution according to the different latitudes of each dam site, the specific method is:

[0026] Firstly, the 24-hour observation values of each reference point are divided into 0-4, 4-8, 8-12, 12-16, 16-20 and 20-0 six time periods;

[0027] Then, the observation value quality of each time period is tested respectively, and the baseline vector under each time period is calculated, and the time period with greater ionospheric delay influence is determined according to the calculation result;

[0028] Finally, the satellite observation values in the time period with greater ionospheric delay influence are removed (the threshold is different under different observation scenes and conditions), and the satellite observation values in other time periods with less ionospheric delay influence are used for GNSS baseline solution;

[0029] Step 23: detecting and repairing frequent small cycle slips, and further optimizing the observation value quality.

[0030] In the above technical solution, in step 23, frequent small cycle slips are detected and repaired, and the observation value quality is further optimized, and the specific method is:

[0031] Firstly, the observation value sequence is fitted by using a time polynomial, the fitting residual is analyzed to find the cycle slip and determine the cycle slip size, the observation value of the fitted phase is subtracted from the calculated value of the phase to obtain a phase bias, when the cycle slip is detected, a bias mark is inserted at the epoch; if no cycle slip is detected, the observation value quality is good, and the step is skipped;

[0032] Then, the double-difference observation values formed by the adjacent three phase observation values are checked to determine whether there is a cycle slip in the double-difference residual sequence; if a cycle slip is found, more double-differences are formed by transforming the satellite and the station to confirm whether there is a cycle slip in the non-difference phase data; this method can correct a cycle or half-cycle small cycle slip; if there is a cycle slip, the cycle slip is repaired; otherwise, the step is skipped;

[0033] By removing the data of the small section with more cycle slip marks, the observation value quality can be significantly improved, and the accuracy and reliability of the GNSS baseline solution are improved. The present application can detect small cycle slips that are easily ignored or cannot be detected by the prior art, which will affect the observation value quality and then affect the solution accuracy; the present application can detect and repair small cycle slips, thereby significantly improving the observation value quality and then improving the accuracy and reliability of the GNSS baseline solution.

[0034] In the above technical solution, in step three, the baseline solution and network adjustment method, specifically:

[0035] Step 31, using measured real-time atmospheric data to correct low-altitude zenith tropospheric delay;

[0036] Firstly, high-precision weather sensors are arranged at the same position of each reference station to collect precise pressure, temperature and humidity information at the station location;

[0037] Secondly, the zenith tropospheric delay (ZTD) of each reference station is estimated by precise point positioning method;

[0038] Then, the new zenith hydrostatic delay (ZHD) is estimated based on Saastamoinen model;

[0039] Finally, the real-time meteorological observation product is brought into the new zenith wet delay (ZWD) model to obtain the precise zenith water vapor distribution characteristics of the station, which significantly improves the accuracy of baseline solution, especially in the vertical direction;

[0040] Step 32, constructing the observation pier thermal expansion model to correct the baseline solution result;

[0041] Firstly, according to the information of the material, structure and height of the observation pier of different reference stations, combined with the measured surface temperature of the observation pier, the displacement of the observation pier caused by the periodic change of temperature is calculated, especially in the vertical direction;

[0042] Then, the thermal expansion displacement model suitable for the station is established, and the model calculation value is added to the baseline solution result as a correction;

[0043] Finally, the period solution in step 22 is used to weaken or even eliminate the high-frequency observation pier thermal expansion displacement caused by high-frequency temperature change by increasing the sampling rate, so as to minimize the influence of temperature change on the accuracy of baseline result based on period solution;

[0044] Step 33, introducing accurate height difference observation value as prior information to increase the redundant observations in network adjustment;

[0045] Step 34, effectively fusing the historical data and measured data under the same observation means, and the data obtained by different observation means.

[0046] In the above technical scheme, in step 31, the process of establishing the new ZWD model is as follows:

[0047] Firstly, the ZWD integral function is defined:

[0048]

[0049] Wherein, R d and R w are the universal dry atmospheric constant and the universal wet atmospheric constant respectively; k1, k2, k3 are the atmospheric refraction index constants; P w represents the atmospheric water vapor pressure; T is the atmospheric temperature; h s is the ground elevation; wherein, the atmospheric temperature can be replaced by the measured atmospheric weighted average temperature T m

[0050] ​

[0051] Thus, equation (1) can be expressed as:

[0052]

[0053] Under the hydrostatic equilibrium condition, the hydrostatic equation can be expressed as:

[0054]

[0055] where P represents the atmospheric pressure; p m represents the atmospheric density; g represents the gravitational acceleration; h is the sea level elevation; the specific humidity q can be generally expressed as a function of the liquid water density p v and the atmospheric density p m :

[0056]

[0057] The liquid water density p v can be expressed as:

[0058]

[0059] If the gravitational acceleration is assumed to be a constant g s , equations (3) to (6) can be combined to obtain:

[0060]

[0061] where P s is the atmospheric pressure at the ground surface;

[0062] The functional relationship between the atmospheric specific humidity and the atmospheric pressure in the vertical distribution can be expressed as:

[0063]

[0064] where q s is the atmospheric specific humidity at the ground surface; γ is the atmospheric mixing ratio parameter; the atmospheric specific humidity can be expressed as a function of the atmospheric pressure and the atmospheric water vapor pressure:

[0065]

[0066] According to equations (8) and (9), we can obtain:

[0067]

[0068] where represents the water vapor pressure at the ground surface;

[0069] The functional relationship between the atmospheric pressure and the atmospheric temperature in the vertical distribution can be expressed as:

[0070]

[0071] where T s is the surface air temperature; β is the temperature lapse rate;

[0072] Combining equation (10) and equation (11) can obtain:

[0073]

[0074] Bringing equation (12) into equation (7) can obtain:

[0075]

[0076] If the atmospheric mixing ratio γ is known, only the surface atmospheric specific humidity, the atmospheric temperature (estimated T m ) and the atmospheric pressure need to be input to estimate the ZWD, and the three parameters can be accurately obtained by the on-site installed weather sensor; the atmospheric mixing ratio γ is an empirical parameter, and in the ideal state, the value is usually 4, and the γ value of different seasons in this area can be obtained according to the measured data.

[0077] In the above technical solution, in step 32, the process of establishing the thermal expansion displacement TEM model suitable for the station is as follows:

[0078] Since the observation pier is generally a concrete structure with poor heat conduction performance, the internal temperature is relatively constant, but the surface temperature will rapidly rise due to solar exposure and other reasons, causing a temperature difference between the surface and the interior of the concrete, and further causing deformation. For the observation pier of metal material, the surface temperature rising speed is particularly rapid, especially at noon in summer. Assuming that the internal temperature of the observation pier is constant, the average air temperature of the station in a certain period of time (such as a season) is replaced, then the thermal expansion displacement of the observation pier in the vertical direction due to temperature change in the period can be represented as:

[0079]

[0080] wherein α is the linear expansion coefficient corresponding to the material of the observation pier, for example, the concrete is 10×10 -6 / ℃, the cast iron pipe is 12×10 -6 / ℃, the copper rod is 17.5×10 -6 / ℃, and the aluminum mast is 23×10 -6 / ℃; H is the vertical height of the observation pier, in mm; T is the surface temperature of the observation pier, which is approximately replaced by the instantaneous air temperature recorded by the weather sensor.

[0081] In the above technical solution, in step 33, the specific method for increasing the number of redundant observations (the number of redundant observations refers to the observation data of TS01 and TS02) in network adjustment is as follows:

[0082] Firstly, two adjacent and moderately spaced reference stations TN01 and TN02 with stable foundation are selected;

[0083] Secondly, reference stations TS01 and TS02 with an observation pier height of 0 (ground mark) are respectively set several meters away from the two reference stations, and GNSS observation is respectively performed;

[0084] Then, the normal height difference △h1 of the baseline TN01-TS01 and the normal height difference △h2 of the baseline TN02-TS02 are measured by precise leveling method. Under such a short distance, we believe that the normal height difference is basically consistent with the geodetic height difference;

[0085] Finally, in the GNSS data processing stage, the geodetic height differences △h1 and △h2 of the baselines TN01-TS01 and TN02-TS02 are introduced as prior information, and the observation data of TS01 and TS02 are integrated with the original reference network for overall adjustment.

[0086] In the above technical solution, in step 34, the historical data and the measured data under the same observation means, and the data obtained by different observation means are effectively fused, and the specific method is as follows:

[0087] For data fusion of different periods: first, remove the gross errors in the time series of various results, and refer to the processing method of the time series analysis method for the step, divide the observation results in the entire time period into different time series, and respectively use linear fitting to fit the time series in different periods; then, according to the fitting result, calculate the difference between the first and last values of adjacent time series, and add the step of the two time series as a known parameter to the time series function model; finally, apply the function model considering the time series step to all observation data of all periods;

[0088] For observation data fusion between different dam vertical displacement monitoring systems: first, the weight information of each observation result should be determined according to the observation accuracy (i.e. the result error) between different dam vertical displacement monitoring systems; then, the weight is distributed to different observation results; finally, according to the error propagation theory, the new data after fusion is calculated by using the weighted average method.

[0089] In the above technical solution, in step 34, the process of gross error elimination is as follows:

[0090] For an observation data sequence {x1, x2, …x N}, the change characteristics of the sequence data are as follows:

[0091] d j = 2x j -(x j+1 +x j-1 )(j = 2, 3, 4, … N-1)(15)

[0092] Thus, N-2 d j values can be obtained from N observation data j The statistical mean value of the sequence data variation can be calculated from d values

[0093]

[0094]

[0095] According to the ratio of the absolute value of the d j deviation to the mean square deviation, the following can be obtained:

[0096]

[0097] When q j > 3, x j is considered to be a singular value and is discarded; otherwise, it is retained.

[0098] The test statistics show that the efficiency of removing singular values in the time series using the method is more than 96%; after removing the singular values, the risk of mixing rough error into the point deformation information can be effectively prevented, and more reasonable and reliable deformation information can be provided for the dam vertical displacement monitoring system.

[0099] In order to achieve the second object of the application, the technical scheme of the application is as follows: the GNSS-based dam vertical displacement monitoring reference network automatic observation method adopts a GNSS vertical displacement monitoring automatic platform, characterized by comprising a data preprocessing module, a baseline solution module and a baseline network adjustment module.

[0100] The data preprocessing module comprises a data quality analysis module, an observation period selection module and a cycle slip detection and repair module.

[0101] The data quality analysis module is used for quality analysis of the measured data, and the main function is to ensure that the GNSS observation data participating in the solution has good quality and avoid mixing in obvious cycle slip or multipath signals.

[0102] The observation period selection module is used for freely selecting the observation period of each reference station participating in the GNSS network solution, determining the optimal observation period and strategy according to the observation value quality of different observation periods, and avoiding the influence of ionospheric activity in low-latitude areas on the observation value quality.

[0103] The cycle slip detection and repair module can detect and repair small cycle slips less than or equal to one week in observation values, and repair the cycle slips to ensure the quality of observation values participating in GNSS network solution;

[0104] The baseline solution module includes a data reading module, an observation value linearization module, a parameter estimation module, a ambiguity fixing module, a troposphere delay correction module, and an observation pier thermal expansion correction module;

[0105] The data reading module can be used to read observation data of each reference station;

[0106] The observation value linearization module is used to linearize different types of observation values (such as L1, L2, L5 of GPS or B1, B2, B3 of Beidou system) to participate in the establishment and solution of normal equations;

[0107] The parameter estimation module and the ambiguity fixing module are used to estimate each ambiguity parameter, and perform least squares estimation according to prior information and normal equations to obtain the estimated value of each unknown parameter for baseline solution;

[0108] The troposphere delay correction module uses measured air pressure, temperature, humidity and other information to implement real-time correction of zenith troposphere delay to improve the solution accuracy;

[0109] The observation pier thermal expansion correction module uses measured temperature and other information such as observation pier height to implement observation pier thermal expansion displacement correction to improve the solution accuracy;

[0110] The baseline network adjustment module can obtain three-dimensional spatial rectangular coordinates of each station through adjustment by setting an initial point and inputting a baseline. The adjustment results can be transformed into a dam coordinate system through built-in coordinate conversion parameters.

[0111] In the above technical solution, the data preprocessing module performs data preprocessing, including the following steps:

[0112] Firstly, the data quality analysis module completes cycle slip detection and repair and gross error detection of observation data to obtain good quality "clean" GNSS observation data, i.e. each epoch and each satellite has corresponding cycle slip information mark, satellite coordinates and clock error, satellite elevation angle, azimuth angle, station coordinates, etc.

[0113] Secondly, after linear combination and linearization of the observation values, the best reference star is searched in all satellites to perform inter-satellite difference to obtain corresponding double difference observation values;

[0114] Thirdly, each epoch and each satellite is directly composed into a normal equation according to the form of normal equation superposition;

[0115] Finally, the LAMBDA algorithm is used to fix the ambiguity after the parameter estimator solves the floating point solution of the equation, and the final baseline fixed solution is obtained by the back equation method.

[0116] To achieve the goal of simple, fast and efficient data calculation, the normal equation is directly composed in the form of superposition for each star per epoch, avoiding the operation of a large number of zero elements of the matrix, improving the calculation efficiency and reducing the computer memory consumption; the maximum advantage of the least square recursive method is fast, but with the increase of the processing arc segment, the number of parameters in the estimator increases significantly, especially the ambiguity parameter, and the calculation speed will be significantly reduced, therefore, the invalid parameters (including ambiguity parameters) should be pre-eliminated and recovered to improve the program efficiency. After the parameter estimator solves the floating point solution of the equation, the LAMBDA algorithm is used to fix the ambiguity, and the final baseline fixed solution is obtained by the back equation method.

[0117] In the above technical solution, the baseline solving module performs baseline solving, and the specific method is as follows:

[0118] (1) setting baseline solving parameters;

[0119] (2) setting the observation data time range according to the optimal observation time period scheme design;

[0120] (3) setting the data path;

[0121] (4) selecting a list of stations to be solved;

[0122] (5) arranging and combining the baselines according to the stations to be solved;

[0123] (6) decompressing the original compressed file according to the file corresponding rule for a certain baseline, searching for the corresponding observation file, ephemeris file and meteorological file;

[0124] (7) correcting the zenith troposphere delay according to the real-time atmospheric parameters;

[0125] (8) implementing thermal expansion displacement correction by using the thermal expansion model;

[0126] (9) solving all single baselines and recording the baseline solving result information.

[0127] The present application has the following advantages:

[0128] The present invention optimizes the deployment plan, adopts precise dual-frequency Beidou receivers and antenna equipment, standardizes the observation pier burial process, and adopts Beidou-based GNSS precision data solution methods and adjustment strategies. Ultimately, a GNSS-based dam vertical displacement reference network monitoring method for reservoir dam deformation monitoring is formed. This method can significantly improve the solution accuracy and reliability of the dam vertical displacement reference network, especially in mountainous and canyon areas with relatively harsh observation environments. Under the premise of ensuring site selection and construction, correct installation of precision receivers and antenna equipment, fine preprocessing of GNSS data, GNSS precision data solution modeling, and GNSS network adjustment with prior information, the present invention constructs an automated vertical displacement monitoring system based on GNSS technology and develops an automated GNSS vertical displacement monitoring platform. The system can improve the vertical baseline solution accuracy from the current common 3-5mm to 1-2mm, especially for baselines with a length of 1-5 kilometers. In some areas with large elevation differences and poor observation conditions (such as areas with elevation differences exceeding 200 meters and low observation value accuracy due to obstruction), the accuracy can reach 2-3mm, which can meet the needs of intelligent development of hydropower station management. The system overcomes the defect of the existing technology that the vertical baseline solution accuracy in areas with large elevation differences and poor observation conditions can only reach a few centimeters or even decimeters, which cannot meet the needs of intelligent development of hydropower station management. For example, for a baseline of about 2 km in length, the mean error of the elevation difference per kilometer is 0.5-1 mm, which can meet the accuracy requirements for second-class leveling in the "National Specifications for First and Second Class Leveling" (GBT 12897-2006) (accidental mean error 1 mm / km, round-trip height difference discrepancy value 5.6 mm); for a baseline exceeding 3 km, under the premise that the accuracy level of the baseline solution remains unchanged, the mean error of the elevation difference per kilometer can reach 0.33-0.67 mm, which meets the accuracy requirements for first-class leveling in the "National Specifications for First and Second Class Leveling" (GBT 12897-2006) (accidental mean error 0.45 mm / km, round-trip height difference discrepancy value 3.1 mm). BRIEF DESCRIPTION OF THE DRAWINGS

[0129] Figure 1 This is a flow chart of the GNSS-based dam vertical displacement reference network monitoring method of the present invention;

[0130] Figure 2 This is a data preprocessing flow chart of the GNSS vertical displacement monitoring automation platform of the present invention;

[0131] Figure 3 This is a diagram of the baseline solution module of the GNSS vertical displacement monitoring automation platform of the present invention. DETAILED DESCRIPTION

[0132] The embodiments of the present application will be described in detail below with reference to the accompanying drawings, but they do not constitute limitations on the present application, and are only examples. At the same time, the advantages of the present application are made clearer and easier to understand through the description.

[0133] The present application improves the vertical direction resolution accuracy of GNSS baseline from 3-5mm to 1-2mm by optimizing the distribution scheme, adopting precise dual-frequency Beidou receiver and antenna equipment, standardizing the observation pier burying process, and simultaneously adopting GNSS precise data solving method and adjustment strategy based on Beidou technology, meeting the intelligent development needs of hydropower station management; in some cases with good observation conditions, even sub-millimeter level can be achieved, which can replace first / second order leveling, and realize automatic monitoring of dam vertical displacement monitoring system, which is the first in China.

[0134] As can be seen from the drawings: in view of the problems existing in the construction of the current GNSS dam vertical displacement monitoring automation system, the main technical solutions and measures of the present application include the following steps:

[0135] To create good hardware conditions and observation environment for GNSS vertical displacement automation monitoring, mainly including optimizing GNSS network shape design and site selection, adopting high-precision Beidou receiver and antenna instrument equipment, and optimizing observation pier foundation and structure design (such as shown in Figure 1

[0136] Step 1, optimize GNSS network shape design and improve site selection standard;

[0137] In the selection of dam vertical displacement monitoring reference network site, in addition to following the selection principles of monitoring reference network in the current national and industry standards and specifications such as “Concrete Dam Safety Monitoring Technical Standard” (GB_T 51416-2020) and “Earth and Rockfill Dam Safety Monitoring Technical Specification” (SL 551-2012), the reference network points should also be selected as far as possible from open water surface, high-voltage tower and higher place, while ensuring that there are no significant obstructions such as mountains and tall trees within 30 degrees of the zenith direction height angle.

[0138] Too flat GNSS network shape or large difference between reference stations will affect the overall stability and reliability of the GNSS network, resulting in low accuracy of network adjustment results. In addition, since the reference station is generally built near large continuous water surface such as reservoir, the multi-path effect caused by signal reflection will seriously affect the quality of observation values received by the receiver. In addition, if there are significant obstructions near the zenith direction of the reference station, it will also have a certain shielding and refraction effect on the satellite signal, affecting the baseline resolution accuracy. The network shape design and site selection standard in the present method can effectively weaken the influence of multi-path effect on GNSS resolution accuracy.

[0139] ​Step 2, high-precision dual-frequency Beidou receiver and choke coil antenna equipment are used;

[0140] With the continuous development of Beidou industry, the quality and algorithm of domestic Beidou GNSS receiver board, chip and other key components have been significantly improved, which has obvious cost advantage compared with expensive imported equipment. In the selection process of receiver, antenna and other hardware, domestic high-precision dual-frequency GNSS receiver supporting Beidou system is selected to ensure that sufficient number and high-quality distance and phase observation values of Beidou, GPS and other different system satellites can be received as much as possible under the premise of saving cost, which provides conditions for subsequent GNSS data precise solution. In addition, choke coil antenna is used, whose suppression plate can absorb more than 90% of satellite signals reflected by the ground, significantly reducing the error caused by multipath effect and improving the quality and precision of baseline solution.

[0141] After determining the position of reference network points and GNSS equipment, the quality of GNSS observation is tested and evaluated to ensure that the obtained observation values meet the provisions of "Global Navigation Satellite System Continuous Operation Reference Station Network Technical Specification" (GB_T28588-2012) on data efficiency, path effect MP1 and MP2 values and cycle slip ratio parameters. Among them, MP1 and MP2 values should be less than 0.5m and 0.65m respectively. If not, the reasons should be analyzed and the site selection of GNSS receiver and antenna equipment should be optimized or replaced.

[0142] Step 3, ensure the stability of observation pier foundation and optimize the structure design of observation pier and its auxiliary facilities;

[0143] In the process of observation pier structure design, the stability of observation pier foundation should be ensured. In principle, the observation pier should be set on bare bedrock, and the observation pier foundation should be anchored in the bedrock more than 1 meter below with sufficient number of steel bars; if the bedrock is located at a deep position below the ground, steel pipe pile should be used to at least 5 meters below the original soil, and concrete should be filled for fixation. If the observation pier foundation is not stable enough, the displacement caused by shallow surface tectonic movement will be transmitted to the GNSS observation results, which will affect or interfere with the calculation results, especially in the vertical direction, which can reach several mm or even centimeter level in some tectonically active areas.

[0144] In addition, the observation pier should avoid using metal pipes (such as copper, iron, etc.) and should use a centrally symmetric structure. This is because the observation pier will be subject to periodic changes in external temperature, including low-frequency (e.g., seasonal cycles) and high-frequency (e.g., daily cycles) thermal expansion displacements. For metal observation piers, the absolute magnitude of thermal expansion displacement can be more than 1.3 times that of ordinary concrete observation piers. Previous studies have shown that if the observation pier is not centrally symmetric, the daily and nightly periodic changes in temperature will also cause millimeter-level thermal expansion displacement in the horizontal direction, and in high-latitude areas, the vertical direction can exceed 1 mm.

[0145] In GNSS vertical displacement automated monitoring, precise data preprocessing methods are used to weaken the influence of multipath effects, retain high-quality GNSS observations, and determine the optimal observation period.

[0146] Step 4: Weaken the influence of multipath effects and eliminate low-quality satellite observations.

[0147] The choke ring antenna can absorb more than 90% of the satellite signals reflected by the ground, significantly reducing the errors caused by multipath effects and improving the quality and accuracy of baseline solutions. In addition, some stations are located in the field without city power, and solar panels and batteries are used for power supply and lightning protection facilities are installed. The solar panels and other auxiliary facilities should not be higher than the bottom of the GNSS choke ring antenna to further weaken the influence of multipath effects.

[0148] Step 5: Select the optimal observation period for solving according to the latitude of each dam site.

[0149] In the GNSS precise data preprocessing process, the optimal observation period method suitable for different dam locations is determined. Different reference networks have different dam locations, with some located in low-latitude areas with active ionospheric activity (such as Yunnan, Guangxi, Guangdong, Hainan, etc.). Generally, the ionospheric activity in low-latitude areas is more intense between 10 am and 12 pm, which may affect satellite observations and result in low baseline solution accuracy. The specific method is as follows: First, divide the 24-hour observations of each reference point into 0-4, 4-8, 8-12, 12-16, 16-20, and 20-0 time periods. Then, test the quality of observations in each period and calculate the baseline vector in each period. According to the solution results, determine the period with greater ionospheric delay influence. Finally, eliminate the observations in the period with greater ionospheric delay influence and use the observations in other periods for GNSS baseline solution.

[0150] Step 6: Detect and repair frequent small cycles to further optimize observation quality.

[0151] In the GNSS precise data preprocessing process, a new method capable of detecting and repairing frequent small cycle slips is adopted. The idea is to use L1 and L2 double difference fitting method to automatically repair cycle slips. First, the time polynomial is used to fit the observation value sequence, the cycle slip is found by analyzing the fitting residual and the cycle slip size is determined. Usually, the observation value of the fitted phase is subtracted from the calculated value. When a cycle slip is detected, a bias marker is inserted at the epoch. Then, the double difference observation value formed by the adjacent three phase observation values is checked to determine whether there is a cycle slip in the double difference residual sequence. If a cycle slip is found, more double differences are formed by transforming the satellite and station to confirm whether there is a cycle slip in the non-difference phase data. This method can correct small cycle slips of one cycle or half cycle. For dam vertical displacement reference network, the baseline is usually short and there are more obstructions such as mountains and buildings around, which will produce more significant cycle slip effect and affect the baseline solution accuracy. By removing the data of the small section with more cycle slip markers, the observation value quality can be significantly improved, and the accuracy and reliability of GNSS baseline solution can be improved.

[0152] In GNSS vertical displacement automatic monitoring, the models such as troposphere delay correction and thermal expansion effect correction are applied to baseline solution, and prior elevation difference information is introduced in network adjustment to improve the accuracy and reliability of baseline solution.

[0153] Step 7, use the measured real-time atmospheric data to correct the low-altitude zenith troposphere delay;

[0154] In the GNSS precise data solution model and method, the low-altitude troposphere correction method based on the measured meteorological parameters is adopted to weaken the influence of troposphere delay on baseline solution. For the dam reference network with large elevation difference of reference station, the vertical displacement accuracy improvement effect is particularly obvious. First, high-precision meteorological sensors are arranged at the same position of each reference station to collect precise atmospheric pressure, temperature and humidity information at the station location. Second, the zenith total delay (ZTD) of each reference station is estimated by using the precise point positioning method. Then, the zenith dry delay (ZHD) is estimated based on the Saastamoinen model. Finally, the real-time meteorological observation product is brought into the new zenith wet delay (ZWD) model to obtain the precise water vapor distribution characteristics of the station, which significantly improves the accuracy of baseline solution, especially in the vertical direction.

[0155] In step 6 of the above technical solution, the process of establishing ZWD model is as follows:

[0156] First, define the ZWD integral function:

[0157]

[0158] Where, R d and R wrespectively, are the universal dry and wet atmospheric constants; k1, k2, k3 are the atmospheric refractive index constants; P w represents the atmospheric water vapor pressure; T is the atmospheric temperature; h s is the surface elevation. In which, the atmospheric temperature can be obtained by the measured atmospheric weighted mean temperature T m is replaced by:

[0159]

[0160] Therefore, the formula (1) can be expressed as:

[0161]

[0162] Under the condition of hydrostatic equilibrium, the hydrostatic equation can be expressed as:

[0163]

[0164] In which, P represents the atmospheric pressure; ρ m represents the atmospheric density; g represents the gravity acceleration; h is the sea level elevation. The specific humidity q can be generally expressed as the function of the liquid water density ρ v and the atmospheric density ρ m .

[0165]

[0166] The liquid water density ρ v can be expressed as:

[0167]

[0168] If the gravity acceleration is assumed to be a constant g s , the joint formula (3)~(6) can be obtained:

[0169]

[0170] In which, P s is the atmospheric pressure at the surface.

[0171] The function relationship of the atmospheric specific humidity and the atmospheric pressure in the vertical distribution can be expressed as:

[0172]

[0173] In which, q s is the atmospheric specific humidity at the surface; γ is the atmospheric mixing ratio parameter. The atmospheric specific humidity can be expressed as the function of the atmospheric pressure and the atmospheric water vapor pressure:

[0174]

[0175] According to the formula (8) and the formula (9), the following can be obtained:

[0176]

[0177] where, represents the water vapor pressure at the surface.

[0178] The function relationship between the atmospheric pressure and the atmospheric temperature in the vertical distribution can be expressed as:

[0179]

[0180] where, T s is the atmospheric temperature at the surface; and β is the temperature lapse rate.

[0181] The joint formula (10) and formula (11) can be obtained:

[0182]

[0183] The formula (12) is brought into the formula (7) to obtain:

[0184]

[0185] If the atmospheric mixing ratio γ is known, only the surface atmospheric specific humidity, the atmospheric temperature (estimated T m ) and the atmospheric pressure need to be input to estimate the ZWD, and the three parameters can be accurately obtained through the on-site installation of the meteorological sensor. The atmospheric mixing ratio γ is an empirical parameter, and in the ideal state, the value is usually 4, and the γ value of different seasons in this region can be obtained according to the measured data.

[0186] Step 8, constructing an observation tower thermal expansion model to modify the baseline calculation results;

[0187] In the GNSS precise data solving model and method, the observation tower thermal expansion model (TEM) considering different observation tower materials and structures is used for modification, so as to weaken the influence of the thermal expansion displacement of the observation tower of the reference station caused by the short-period change of the temperature on the calculation results. First, according to the information such as the material, structure and height of the observation tower of different reference stations, combined with the measured surface temperature of the observation tower, the displacement of the observation tower caused by the periodic change of the temperature is calculated, especially in the vertical direction. Then, the thermal expansion displacement model suitable for the station is established, and the model calculation value is added to the baseline calculation results as a correction. Finally, the time interval solution in step 5 is used to weaken or even eliminate the high-frequency thermal expansion displacement of the observation tower caused by the high-frequency temperature change by increasing the sampling rate, so as to minimize the influence of the temperature change on the baseline result precision based on the time interval solution.

[0188] In step 7 of the above technical solution, the process of establishing the TEM model is as follows:

[0189] Since the observation pier is generally a concrete structure with poor thermal conductivity, the internal temperature is relatively constant, but the surface temperature will rapidly rise due to solar radiation and other reasons, causing a temperature difference between the surface and the interior of the concrete, and further causing deformation. For observation piers made of metal, the surface temperature rises particularly rapidly, especially at noon in summer. Assuming that the internal temperature of the observation pier is constant, the average air temperature at the location of the reference station during a certain period (such as a season) is taken as the internal temperature of the observation pier Alternatively, the thermal expansion displacement of the observation pier in the vertical direction due to temperature changes during the period can be represented as:

[0190]

[0191] where α is the linear expansion coefficient corresponding to the material of the observation pier, for example, 10×10 -6 / ℃ for concrete, 12×10 -6 / ℃ for cast iron pipe, 17.5×10 -6 / ℃ for copper rod, and 23×10 -6 / ℃ for aluminum mast. H is the vertical height of the observation pier, in mm. T is the surface temperature of the observation pier, which is approximately replaced by the instantaneous air temperature recorded by the weather sensor.

[0192] Step 9: Introduce precise height difference observation values as prior information to increase the number of redundant observations in network adjustment;

[0193] In the GNSS precise data solution model and method, the network layout scheme is optimized, and the fixed height difference short baseline method is added at the conditional stations to improve the overall solution accuracy of the reference network. The specific method is as follows: first, select two adjacent and moderately spaced reference stations TN01 and TN02 with stable foundations; second, set up reference stations TS01 and TS02 with an observation pier height of 0 (ground mark) several meters away from the two reference stations, and perform GNSS observation respectively; then, measure the normal height difference △h1 of baseline TN01-TS01 and the normal height difference △h2 of baseline TN02-TS02. Under such a short distance, we believe that the normal height difference is basically the same as the geodetic height difference; finally, introduce the geodetic height differences △h1 and △h2 of baselines TN01-TS01 and TN02-TS02 as prior information in the GNSS data processing stage, and perform overall adjustment of the observation data of TS01 and TS02 together with the original reference network.

[0194] This step not only introduces the more accurate leveling measurement results as prior information into the network adjustment to improve the accuracy of the adjustment results, but also increases the corresponding ultra-short baseline observation values as redundant observations, which can effectively enhance the reliability of the adjustment results.

[0195] Step 10, effectively fuse the historical data and the measured data under the same observation means, and the data obtained by different observation means;

[0196] The data of the new and old observation systems are integrated, and the data processing and fusion methods for different periods and different types of vertical displacement reference network are adopted to fuse the historical data and the measured data, the leveling measurement data and the GNSS measurement data. For the data of different periods, first, the gross errors in the time series of various results are removed, the observation results in the whole time period are divided into different time series by referring to the processing method of the jump in the time series analysis method, and the linear fitting method is used to fit the time series of different periods; then, according to the fitting results, the difference between the first and last values of adjacent time series is calculated, and the jump of the two time series is added to the time series function model as a known parameter; finally, the function model considering the jump of the time series is applied to the observation data of all periods. For the observation data between different systems, first, the weight information of the observation results of each system should be determined according to the observation accuracy (i.e. the result error) between different systems; then, the weight is distributed to different observation results; finally, the new fused data is calculated by using the weighted average method according to the error propagation theory.

[0197] In step 10 of the above technical solution, the process of removing gross errors is as follows:

[0198] For the observation data sequence {x1, x2,…x N}, the change characteristics of the sequence data are described as

[0199] d j =2x j -(x j+1 +x j-1 )(j=2,3,4,…N-1)(15)

[0200] In this way, N-2 d j values can be obtained from N observation data. j At this time, the statistical mean and the mean square error of the sequence data change can be calculated from the d j values.

[0201]

[0202]

[0203] According to the ratio of the absolute value of the d j deviation to the mean square error, the q j value can be obtained.

[0204]

[0205] When q j > 3, xj is a singular value, which is discarded.

[0206] The test statistics show that the efficiency of removing singular values in the time series by using the method is more than 96%. After removing the singular values, the rough error can be effectively prevented from mixing into the point deformation information, the risk can be avoided, and more reasonable and reliable deformation information can be provided for the system.

[0207] On the basis of steps 1 to 10, the application establishes a set of GNSS technology-based vertical displacement monitoring automation system and develops a GNSS vertical displacement monitoring automation platform.

[0208] The GNSS vertical displacement monitoring automation platform adopts B / S structure and is developed based on.NET technology. The GNSS vertical displacement monitoring automation platform mainly includes a data preprocessing module, a baseline solution module and a baseline network adjustment module.

[0209] 1) The data preprocessing module includes a data quality analysis module, an observation period selection module and a cycle slip detection and repair module;

[0210] (1) The data quality analysis module can perform quality analysis on the measured data, and the main function is to ensure that the GNSS observation data participating in the solution has good quality, so as to avoid mixing in obvious cycle slip or multipath signals;

[0211] (2) The observation period selection module can freely select the observation period of each reference station participating in the GNSS network solution, determine the optimal observation period and strategy according to the quality of the observation values of different observation periods, and avoid the influence of ionospheric activity in low-latitude areas on the quality of observation values;

[0212] (3) The cycle slip detection and repair module can detect and repair small cycle slips less than or equal to one week in observation values, and repair them to ensure the quality of observation values participating in the GNSS network solution.

[0213] The specific process of data preprocessing performed by the data preprocessing module is shown in the attached Figure 2As shown, including the following steps: first, the data quality control module completes the cycle slip detection and repair and gross error detection of observation data, and obtains "clean" GNSS observation data, that is, each epoch each star has corresponding cycle slip information mark, satellite coordinates and clock error, satellite elevation angle, azimuth angle, station coordinates and other information. After linear combination and linearization of the observation values, the best reference star is searched in all satellites to perform inter-satellite difference, and the corresponding double difference observation values are obtained. In order to realize the goal of simple, fast and efficient data calculation, the normal equation is directly composed in the form of normal equation superposition for each epoch and each star, avoiding the operation of a large number of zero elements of the matrix, improving the calculation efficiency and reducing the computer memory consumption. The biggest advantage of the least square recursive method is fast, but as the number of processing arcs increases, the number of parameters in the estimator increases significantly, especially the ambiguity parameters, and the calculation speed will be significantly reduced, so the invalid parameters (including ambiguity parameters) should be pre-eliminated and recovered to improve the program efficiency. After the floating point solution of the equation is solved by the parameter estimator, the LAMBDA algorithm is used to fix the ambiguity, and the back propagation equation is used to obtain the final baseline fixed solution.

[0214] 2) The baseline solution module includes a data reading module, an observation value linearization module, a parameter estimation module, and an ambiguity fixing module, a troposphere delay correction module, and an observation pier thermal expansion correction module (as shown in Figure 3

[0215] (1) The data reading module can read the observation data of each reference station;

[0216] (2) The observation value linearization module linearizes different types of observation values (such as GPS L1, L2, L5 or Beidou system B1, B2, B3 frequency) after linearization, and participates in the establishment and solution of the normal equation;

[0217] (3) The parameter estimation module and the ambiguity fixing module estimate each ambiguity parameter, and perform least square estimation according to the prior information and the normal equation to obtain the estimated value of each unknown parameter for baseline solution;

[0218] (4) The troposphere delay correction module uses the measured air pressure, air temperature, humidity and other information to implement real-time correction of zenith troposphere delay, and improves the solution accuracy;

[0219] (5) The observation pier thermal expansion correction module uses the measured temperature and combines with the observation pier height and other information to implement the observation pier thermal expansion displacement correction, and improves the solution accuracy;

[0220] The overall process of the baseline solution module for baseline solution is as follows:

[0221] (1) Set the baseline solution parameters;

[0222] ​(2) According to the optimal observation period scheme design, set the observation data time range;

[0223] (3) Set the data path;

[0224] (4) Select the list of stations to be solved;

[0225] (5) According to the station to be solved baseline, arrange combination;

[0226] (6) According to the file corresponding rule, decompress the original compressed file, search for the corresponding observation file, ephemeris file, meteorological file;

[0227] (7) According to the real-time atmospheric parameter, carry out zenith troposphere delay correction;

[0228] (8) Use thermal expansion model to implement thermal expansion displacement correction;

[0229] (9) Solve all single baseline, record the baseline solution result information.

[0230] 3) Baseline network adjustment module

[0231] By setting the starting point and inputting the baseline, the three-dimensional space rectangular coordinates of each station can be obtained through adjustment. Through the built-in coordinate conversion parameters, the benchmark transformation of the adjustment result can be carried out to obtain the dam coordinate system results. In addition, according to the prior elevation information, the overall network adjustment can be carried out again to improve the reliability of the adjustment result.

[0232] The other parts not mentioned belong to the prior art.

Claims

1. A GNSS-based automated observation method for a dam vertical displacement monitoring benchmark network, characterized by: The following steps are included: Step 1: Site survey, instrument and observation pier selection; Optimizing GNSS grid design and point selection, adopting high-precision BeiDou receivers and antenna equipment, and optimizing the foundation and structural design of observation piers will create favorable hardware conditions and an observation environment for automated GNSS vertical displacement monitoring. Step 2: GNSS data preprocessing and solution optimization; In the automated monitoring of GNSS vertical displacement, precise data preprocessing methods are used to reduce the impact of multipath effects, retain high-quality GNSS observations, determine the optimal observation period, and correct frequent small cycle slips. Step 3: Baseline solution and network adjustment method; In the automated monitoring of GNSS vertical displacement, the low-altitude zenith tropospheric delay correction and thermal expansion effect correction models are applied to the baseline solution. Priori height difference information is introduced into the network adjustment to improve the accuracy and reliability of the baseline solution used to calculate dam displacement. In step 3, the baseline solution and network adjustment methods are as follows: Step 31, using the measured real-time atmospheric data to perform low-altitude zenith tropospheric delay correction; First, high-precision meteorological sensors are co-located at each reference station to collect precise air pressure, temperature, and humidity information at the station location; Secondly, the total tropospheric zenith delay of each reference station is estimated using the precise point positioning method; Then, a new zenith dry delay is estimated based on the Saastamoinen model; Finally, the real-time meteorological observation products are introduced into the new zenith wet delay model to obtain the precise water vapor distribution characteristics at the zenith of the observation station. Step 32, constructing a correction model for thermal expansion effect of the observation pier to revise the baseline solution result; First, based on the material, structure, and height information of the observation piers at different reference stations and the measured surface temperature of the observation piers, the displacement of the observation piers caused by the periodic temperature change is calculated. Then, a thermal expansion displacement model suitable for the measuring station is established, and the model calculation value is added to the baseline solution result as a correction number; Finally, using the time period solution in step 22, the high-frequency thermal expansion displacement of the observed pier caused by high-frequency temperature changes is weakened or even eliminated by increasing the sampling rate, thereby minimizing the impact of temperature changes on the accuracy of the baseline results based on the time period solution. Step 33: introduce accurate height difference observations as prior information to increase the number of redundant observations in the network adjustment; Step 34: effectively integrate the historical data and measured data obtained under the same observation method and the data obtained by different observation methods.

2. The GNSS-based automated observation method for dam vertical displacement monitoring reference network according to claim 1 is characterized by: In step one, site selection, instrument and observation pier selection include the following steps: Step 11: Optimize GNSS grid design and improve point selection standards; In selecting the sites for the dam vertical displacement monitoring benchmark network, in addition to adhering to the principles for selecting monitoring benchmark networks in the current national and industry standards and specifications, GB_T 51416-2020 "Technical Standard for Safety Monitoring of Concrete Dams" and SL 551-2012 "Technical Specification for Safety Monitoring of Earth-Rock Dams," the benchmark network sites were also located at higher locations away from open water and high-voltage towers, while ensuring that there were no significant obstructions within a 30-degree zenith elevation angle. Step 12: Use a high-precision dual-frequency Beidou receiver and choke antenna equipment; Step 13: Ensure the stability of the observation pier foundation and optimize the structural design of the observation pier and ancillary facilities; The observation pier is set on the exposed bedrock, and the foundation of the observation pier is anchored at least 1 meter below the bedrock with steel bars; if the bedrock is located deeper below the ground, steel pipe piles are used to at least 5 meters below the original soil and filled with concrete for fixation.

3. The GNSS-based automated observation method for dam vertical displacement monitoring reference network according to claim 1 or 2, characterized in that: In step 2, GNSS data preprocessing and solution optimization are performed. The specific methods are as follows: Step 21: Reduce the impact of multipath effects; Choke ring antennas are used, and their path suppression plates absorb more than 90% of satellite signals reflected by the ground, significantly reducing errors caused by multipath effects. Ensure that the solar panels and other ancillary facilities are no higher than the bottom of the GNSS choke antenna to further reduce the impact of multipath effects; Step 22: Eliminate low-quality satellite observations and select the optimal observation period for calculation based on the latitude of each dam site. The specific method is as follows: First, the 24-hour observation values ​​of each benchmark point are divided into six periods: 0:00 to 4:00, 4:00 to 8:00, 8:00 to 12:00, 12:00 to 16:00, 16:00 to 20:00, and 20:00 to 0:00; Then, the quality of the observations in each period is tested, and the baseline vector in each period is calculated. The period with the greater influence of ionospheric delay is determined based on the solution results. Finally, the satellite observations during the period with greater ionospheric delay are eliminated, and the satellite observations during other periods with less ionospheric delay are used to calculate the GNSS baseline. Step 23: Detect and repair frequent small cycle slips to further optimize the observation quality.

4. The GNSS-based automated observation method for dam vertical displacement monitoring reference network according to claim 3 is characterized by: In step 23, frequent small cycle slips are detected and repaired to further optimize the observation quality. The specific method is as follows: First, a time polynomial is used to fit the observation sequence. The fitting residuals are analyzed to detect cycle slips and determine their magnitude. The calculated phase value is subtracted from the observed value of the fitted phase to obtain the phase deviation. When a cycle slip is detected, a deviation marker is inserted at that epoch. If no cycle slip is detected, it indicates that the observation quality is good and this step is skipped. Then, the double difference observations composed of three adjacent phase observations are checked to determine whether there is a cycle slip in the double difference residual sequence; If a cycle slip is found, more double differences are generated by changing the satellite and the station, and the double difference phase data is checked to see if a cycle slip exists. If so, the cycle slip is repaired; otherwise, this step is skipped. By eliminating small segments of data with more cycle slip markers, the quality of observations can be significantly improved.

5. The GNSS-based automated observation method for dam vertical displacement monitoring reference network according to claim 1 is characterized by: In step 31, the process of establishing a new zenith wet delay model is as follows: First, define the zenith wet delay integral function: Among them, R d and R w are the universal dry atmosphere constant and the universal wet atmosphere constant respectively; k1, k2, k3 are the atmospheric refractive index constants; P w represents atmospheric water vapor pressure; T is atmospheric temperature; h s is the surface elevation; the atmospheric temperature is the measured atmospheric weighted average temperature T m replace: Therefore, formula (1) is expressed as: Under the condition of hydrostatic equilibrium, the hydrostatic equation can be expressed as: Where P represents the atmospheric pressure; ρ m represents the atmospheric density; g represents the acceleration due to gravity; h represents the sea level elevation; and specific humidity q represents the density of liquid water ρ. v and atmospheric density ρ m Function: Liquid water density ρ v Expressed as: If we assume that the acceleration due to gravity is a constant g s , combining formulas (3) to (6) to obtain: Among them, P s is the atmospheric pressure at the surface; The functional relationship between atmospheric specific humidity and atmospheric pressure in vertical distribution is expressed as: Among them, q s is the atmospheric specific humidity at the surface; γ is the atmospheric mixing ratio parameter; atmospheric specific humidity is expressed as a function of atmospheric pressure and atmospheric water vapor pressure: in, Indicates the water vapor pressure at the surface; The functional relationship between atmospheric pressure and atmospheric temperature in vertical distribution is expressed as: Among them, T s is the atmospheric temperature at the surface; β is the temperature lapse rate; Combining formula (10) and formula (11) we get: Substituting formula (12) into formula (7) yields: If the atmospheric mixing ratio γ is known, ZWD can be estimated by inputting the surface atmospheric specific humidity, atmospheric temperature, and atmospheric pressure. These three parameters are accurately obtained through meteorological sensors installed on site; the atmospheric mixing ratio γ is an empirical parameter.

6. The GNSS-based automated observation method for a dam vertical displacement monitoring reference network according to claim 5, characterized in that: In step 32, the process of establishing the thermal expansion displacement model applicable to the measuring station is as follows: Assuming that the temperature inside the observation tower is constant, the average temperature of the location of the base station during a certain period of time is used. Instead, the thermal expansion displacement of the observation pier in the vertical direction due to temperature change during this period of time can be expressed as: Where α is the linear expansion coefficient corresponding to the material of the observation pier, H is the vertical height of the observation pier, in mm; T is the surface temperature of the observation pier, which is approximately replaced by the instantaneous air temperature recorded by the meteorological sensor.

7. The GNSS-based automated observation method for a dam vertical displacement monitoring reference network according to claim 6, characterized in that: In step 33, the specific method for increasing the number of redundant observations in the network adjustment is as follows: First, two adjacent benchmark stations TN01 and TN02 with moderate spacing and relatively stable foundations were selected; Secondly, the reference stations TS01 and TS02 with an observation pier height of 0 were set up a few meters away from the two reference station observation piers, and GNSS observations were carried out respectively; Then, the precision leveling method was used to measure the normal height difference △h1 of the baseline TN01-TS01 and the normal height difference △h2 of the baseline TN02-TS02. At such a short distance, it is believed that the normal height difference is basically consistent with the geoid height difference; Finally, in the GNSS data processing stage, the geodetic height differences △h1 and △h2 of the baselines TN01-TS01 and TN02-TS02 are introduced as prior information, and the observation data of TS01 and TS02 are overall adjusted together with the original reference network.

8. The GNSS-based automated observation method for a dam vertical displacement monitoring reference network according to claim 7, characterized in that: In step 34, historical data and measured data under the same observation method and data obtained by different observation methods are effectively integrated. The specific method is as follows: For data fusion of different periods: First, the gross errors in the time series of various results are eliminated. By referring to the treatment method of steps in time series analysis methods, the observation results in the entire time period are divided into different time series. The time series of different periods are fitted by linear fitting respectively; then, based on the fitting results, the difference between the first and last values ​​of adjacent time series is calculated, and the steps of these two time series are added as known parameters to the time series function model to form a function model that takes into account the time series steps; finally, the function model that takes into account the time series steps is applied to the observation data of all periods; For the fusion of observation data between different dam vertical displacement monitoring systems: First, the weight information of each observation result is determined according to the observation accuracy between different dam vertical displacement monitoring systems; Then, weights are assigned to different observation results; finally, the weighted average method is used according to the error propagation theory to calculate the new fused data.

9. The GNSS-based automated observation method for a dam vertical displacement monitoring reference network according to claim 8, characterized in that: In step 34, the process of eliminating gross errors is as follows: For the observation data sequence {x1,x2,…x N }, describing the change characteristics of the sequence data as follows: d j =2x j -(x j+1 +x j-1 ), where (j=2,3,4,…N-1)(15) In this way, N-2 d values ​​are obtained from N observation data. j , at this time, by d j Calculate the statistical mean of the change in sequence data and mean square error According to d j The ratio of the absolute value of the deviation to the mean square error is obtained: When q j >3, it is considered that x j If it is a singular value, it will be discarded; otherwise, it will be retained.

10. The GNSS vertical displacement monitoring automation platform used in the GNSS-based dam vertical displacement monitoring reference network automated observation method according to any one of claims 1 to 9, characterized in that: It includes data preprocessing module, baseline solution module and baseline network adjustment module; The data preprocessing module includes a data quality analysis module, an observation period selection module, and a cycle slip detection and repair module; The data quality analysis module is used to perform quality analysis on the measured data to ensure the quality of the GNSS observation data involved in the solution is good and to avoid mixing with signals with obvious cycle slips or multipath; The observation period selection module is used to freely select the observation period of each base station participating in the GNSS network solution. According to the quality of observation values ​​in different observation periods, the optimal observation period and strategy are determined to avoid the impact of active ionospheric activity in low latitudes on the quality of observation values. The cycle slip detection and repair module detects and repairs small cycle slips of less than or equal to one cycle in the observation value, and repairs them to ensure the quality of the observation value participating in the GNSS network solution; The baseline solution module includes a data reading module, an observation linearization module, a parameter estimation module, an ambiguity fixation module, a tropospheric delay correction module, and an observation pier thermal expansion correction module; Data reading module, used to read the observation data of each reference station; The observation linearization module is used to linearize different types of observations and then participate in the establishment and solution of the normal equation; The parameter estimation module and the ambiguity fixation module are used to estimate the ambiguity parameters and perform least squares estimation based on prior information and the normal equation to obtain the estimated values ​​of the unknown parameters and perform baseline solution. The tropospheric delay correction module uses measured air pressure, temperature, and humidity information to implement real-time correction of zenith tropospheric delay to improve solution accuracy; The observation pier thermal expansion correction module uses the measured temperature and the observation pier height information to implement the observation pier thermal expansion displacement correction to improve the solution accuracy; The baseline network adjustment module sets the starting point and inputs the baseline to obtain the three-dimensional rectangular coordinates of each measuring station through adjustment; the adjustment results are subjected to benchmark transformation through the built-in coordinate conversion parameters to obtain the dam coordinate system results.

11. The GNSS vertical displacement monitoring automation platform according to claim 10, characterized in that: The data preprocessing module performs data preprocessing, including the following steps: First, the data quality analysis module completes cycle slip detection and repair as well as gross error detection of observation data to obtain high-quality GNSS observation data. Secondly, after linearly combining and linearizing the observation values, the best reference satellite is found among all satellites to perform inter-satellite difference and obtain the corresponding double-difference observation values; Thirdly, the normal equations for each star in each epoch are directly composed in the form of superposition of normal equations; Finally, the floating-point solution of the equation is solved by the parameter estimator, and the LAMBDA algorithm is used to fix the ambiguity, and the final baseline fixed solution is obtained by the back-banding method equation.

12. The GNSS vertical displacement monitoring automation platform according to claim 11, characterized in that: The baseline calculation module performs baseline calculation. The specific method is as follows: (1) Set baseline solution parameters; (2) Design the optimal observation period and set the observation data time range; (3) Set the data path; (4) Select the list of stations to be solved; (5) Arrange and combine the baselines according to the stations for which the baselines need to be solved; (6) For a certain baseline, decompress the original compressed file according to the file correspondence rule and search for the corresponding observation file, ephemeris file, and meteorological file; (7) Perform zenith tropospheric delay correction based on real-time atmospheric parameters; (8) Implement thermal expansion displacement correction using thermal expansion model; (9) Solve all single baselines and record the baseline solution result information.

Images