CN-122017902-A - GNSS real-time dynamic calculation system and method for automatic monitoring of dam deformation

CN122017902ACN 122017902 ACN122017902 ACN 122017902ACN-122017902-A

Abstract

The invention relates to the technical field of sensor networks and state sensing, in particular to a GNSS real-time dynamic calculation system and method for dam deformation automatic monitoring; the method comprises the steps of monitoring node running states through dynamic perception of GNSS, automatically forming and updating a resolving reference role, constructing a state vector based on multidimensional state information, carrying out normalization and reliability assessment, adaptively generating a reference node candidate set, carrying out residual quantization, environment disturbance correction and self-correction through a reference node time sequence evolution model, integrating a self-correction reference field and observation data through cross-epoch coupling resolving, realizing real-time continuous and steady displacement resolving and generating node reliability, coupling resolving results with dam structure mechanical constraint, carrying out closed loop verification and abnormal node dynamic correction, and constructing a multi-model fusion prediction system to realize future displacement prediction, risk assessment and grading early warning. The invention improves the automation, the intellectualization and the stability of the dam deformation monitoring.

Inventors

XIONG JING
TIAN PENG
SONG YUANWEI
WANG DONGCAI
MA JINPING
LI XIANQI
ZHANG JIANGUO
LIU ZHAOYOU
REN SHIYU
LI YUNLONG
WU JIE
LIN FENG
HE LIHUA
LI HAIBO
YANG ZHIHU
LUO KEJUN
LUO JIANZHE
LIU JIN

Assignees

云南华电金沙江中游水电开发有限公司梨园发电分公司
中国电建集团贵阳勘测设计研究院有限公司

Dates

Publication Date: 20260512
Application Date: 20260227

Claims (10)

1. The GNSS real-time dynamic resolving method for automatically monitoring the deformation of the dam is characterized by comprising the following specific implementation steps: s1, dynamically sensing the running state and resolving behavior of each monitoring node in a GNSS monitoring network, automatically forming and dynamically updating resolving reference roles based on the observation continuity, resolving stability and environment response multidimensional state information of the nodes; S2, constructing a reference node time sequence evolution model, carrying out inter-node relative coordinate residual error statistics and quantification on a selected reference node set, introducing an environmental disturbance factor for correction, generating a self-correction coordinate increment to update a reference node coordinate, and carrying out overall reliability assessment on an updated reference field; S3, performing cross-epoch coupling resolving, constructing a multi-epoch state vector, performing state prediction by utilizing a reference field evolution trend, coupling real-time observation node data with self-correction reference node coordinates to construct a dynamic gain matrix, updating the node state vector, calculating a resolving residual error, and generating node resolving credibility; S4, coupling a multi-epoch GNSS (Global navigation satellite System) resolving result with dam structure mechanics constraint, constructing a structure mechanics constraint model, calculating displacement residual errors, identifying abnormal nodes by combining node credibility, carrying out weighted least square closed loop correction, dynamically updating node weights, and evaluating the whole network correction displacement consistency; And S5, based on the corrected node displacement sequence, the historical trend data and the environmental disturbance information, constructing a multi-model fusion prediction system to realize future displacement prediction, calculating a node future displacement super-threshold risk index, weighting according to node credibility to generate a whole network risk index, dividing the early warning level according to the index, and feeding back a prediction result to a node management and resolving process in a closed loop.
2. The method for dynamically calculating the GNSS real-time data of automatic dam deformation monitoring according to claim 1, wherein in step S1, the dynamically sensing the operation states and the calculation behaviors of each monitoring node in the GNSS monitoring network specifically comprises: Continuously collecting and quantifying observation integrity, signal quality, calculation result change and environment disturbance response of each GNSS monitoring node, and constructing a unified node operation state vector; Carrying out standardization and trend smoothing treatment on each state component in the node original state vector to obtain a state characteristic value under a unified scale, and introducing a solution continuous consistency characteristic to construct a solution stability core index; Constructing a node comprehensive credibility function based on the multidimensional state characteristics, fusing the multidimensional state characteristics into uniform credibility evaluation, and adaptively generating a reference node candidate set according to the overall credibility distribution of the network; and continuously monitoring node reliability change, and triggering automatic adjustment and update of the resolving role when the reference node reliability is lower than a minimum threshold value or the non-reference node reliability is higher than the average reliability of the reference node.
3. The method for dynamically resolving a dam deformation automatically monitored GNSS in real time according to claim 2, wherein the node operation state vector is specifically: The node running state vector comprises an observation continuity index, an observation quality stability index, a short-term consistency index of a solution result, an environment disturbance response index and a historical role reliability index; The observation continuity index reflects the capability of the node to stably provide effective GNSS original observation data, the observation quality stability index is derived from comprehensive evaluation of carrier phase noise, multipath influence and signal fluctuation conditions, the short-term consistency index of a resolving result is used for measuring the stability of the result change of the node in the continuous epoch resolving process, the environmental disturbance response index is used for describing the sensitivity degree of the node to external environmental factors, and the historical role reliability index is used for describing the contribution degree of the node to the overall resolving stability when the node is used as a reference node in the historical operation process.
4. The method for GNSS real-time dynamic solution for dam deformation automatic monitoring according to claim 3, wherein in step S2, constructing a reference node time-sequence evolution model specifically includes: calculating relative coordinate residual errors among nodes in a reference node set in a sliding time window, and carrying out time weighted sliding average on a residual error sequence to quantify the tiny evolution trend of a reference field; introducing an environment disturbance factor matrix containing temperature, wind speed, humidity and shielding type environment disturbance variables, and coupling an environment disturbance correction vector with a reference node residual sequence to remove unstructured offset; generating a self-correction coordinate increment for each reference node based on the corrected residual sequence, and updating the reference node coordinates accordingly; And carrying out overall reliability evaluation on the updated reference field, and triggering reference node reselection or weight redistribution when the overall evolution reliability of the reference field obtained by evaluation is lower than a set threshold value.
5. The method for dynamically resolving a dam deformation automatically monitored GNSS in real time according to claim 4, wherein in step S3, the cross-epoch coupling resolving specifically comprises: Based on the self-correction reference node set and time sequence evolution parameters thereof, organizing coordinates of all monitoring nodes into multi-epoch state vectors in time sequence; Constructing a cross-epoch prediction model, and using a state transition matrix to use the history displacement and the evolution trend of the reference field for the current epoch state prediction; Constructing an observation equation, coupling real-time observation node data with self-correction reference node coordinates, and constructing a dynamic coupling gain matrix; the observed noise covariance matrix is dynamically adjusted according to node historical calculation reliability; updating the node state vector by using the gain matrix, calculating a solution residual sequence, and performing time weighted analysis on the residual sequence to evaluate node solution stability; and generating the calculated credibility of each node according to residual evolution, and feeding back the calculated credibility to be used for dynamically adjusting the distribution of the reference node and the observation node.
6. The method for dynamically resolving a dam deformation automatically monitored GNSS in real time according to claim 5, wherein in step S4, coupling the multi-epoch GNSS resolving result with the dam structural mechanical constraint specifically comprises: Decomposing a dam into a plurality of structural units, defining an elastic mechanical constraint equation to correlate node displacement vectors with structural constraint conditions, and introducing different weight coefficients to different dam segments and key monitoring points to form weighted structural constraint; Calculating residual errors between the calculated displacement of each node and the constraint of the weighted structure, and carrying out weighted analysis by combining the calculated reliability of the nodes so as to identify abnormal nodes; Performing closed-loop correction on the identified abnormal node by adopting a weighted least square method, wherein the correction weight of the node is calculated by combining the node reliability and residual error size; And calculating a full-network correction displacement consistency index, triggering the updating of the reference nodes and the redistribution of the node weights if the index is lower than a set threshold, and taking the corrected displacement as the input of a dynamic evolution model of the reference field.
7. The method for automatically monitoring the deformation of the dam by using the GNSS real-time dynamic calculation method according to claim 6, wherein in the step S5, the construction of the multi-model fusion prediction system specifically comprises the following steps: Collecting the whole network correction displacement sequence and node credibility to form a sliding time window historical data set, extracting key characteristic indexes of short-term displacement speed, displacement acceleration and residual variance from the historical data set, and weighting the characteristics by combining the node credibility; Constructing a multi-model prediction system comprising a weighted ARMA model, a structural mechanical coupling model and a machine learning nonlinear model for each node; the weighted ARMA model predicts short-term period fluctuation by utilizing node historical displacement and residual error in a sliding time window; the structural mechanical coupling model predicts a long-term trend based on structural constraints and environmental loads; machine learning nonlinear models capture nonlinear historic laws using LSTM; And dynamically distributing weights according to the prediction results of the three types of models and fusing the prediction results according to the historical prediction errors to obtain a node fusion prediction displacement result.
8. The method for automatically monitoring dam deformation in real time and dynamically resolving by GNSS according to claim 7, wherein the machine learning nonlinear model is a long-term and short-term memory network model for capturing nonlinear and complex time-dependent rules in the node displacement history data.
9. The method for automatically monitoring dam deformation by GNSS real-time dynamic calculation according to claim 8, wherein the step of classifying the pre-warning level according to the index comprises the following steps: Calculating a node future displacement super-threshold risk index according to the fusion prediction displacement result of the node and the corresponding safety displacement threshold determined by dam body design, historical deformation statistics and material characteristics; Weighting and averaging risk indexes of all nodes of the whole network according to the node reliability to generate a risk index of the whole network; And dividing green, yellow, orange and red early warning grades according to the numerical range of the whole-network risk index.
10. A GNSS real-time dynamic solution system for automatic monitoring of dam deformation, configured to perform a GNSS real-time dynamic solution method for automatic monitoring of dam deformation according to any of claims 1to 9, comprising: the node monitoring and state sensing module is used for sensing the running state and the data quality of each GNSS monitoring node in real time, and realizing node reliability calculation and dynamic allocation and adjustment of reference roles based on multidimensional state evaluation; the multi-epoch real-time dynamic resolving module is used for performing cross-epoch coupling resolving based on the self-correction reference field and the observation data, updating the node displacement state and generating resolving credibility; The structure consistency check and constraint fusion module is used for integrating a dam structure mechanical constraint model, carrying out consistency check on the calculated displacement, and identifying and correcting abnormal node displacement; The future displacement prediction and risk assessment module is used for predicting the node displacement based on the historical data and the multi-model fusion strategy, calculating the risk index and realizing hierarchical early warning; and the closed-loop feedback and intelligent regulation module is used for feeding back the prediction and risk assessment results to the modules, dynamically regulating the monitoring strategy and generating a monitoring report and a regulation instruction.

Description

GNSS real-time dynamic calculation system and method for automatic monitoring of dam deformation Technical Field The invention relates to the technical field of sensor networks and state sensing, in particular to a GNSS real-time dynamic resolving system and method for dam deformation automatic monitoring. Background Along with the continuous improvement of the positioning precision and reliability of the global navigation satellite system, the global navigation satellite system is increasingly widely applied to the automatic monitoring of the deformation of heavy engineering structures such as dams and the like, the current monitoring technology can realize the continuous observation of millimeter-level displacement, and engineering practice brings higher requirements on the long-term stability, environmental adaptability and result reliability of the monitoring system. The invention patent of China with the bulletin number of CN116123982B discloses an automatic observation method for a dam vertical displacement monitoring reference network based on GNSS, which comprises the following steps of point location investigation and selection of instruments and observation piers, the second step of GNSS data preprocessing and scheme optimization, the determination of an optimal observation period in the automatic GNSS vertical displacement monitoring and the repair of frequent small cycle hops, the third step of baseline calculation and net adjustment method, the application of models for correcting troposphere delay and thermal expansion effect in baseline calculation in the automatic GNSS vertical displacement monitoring, and the introduction of priori altitude difference information in net adjustment, thereby improving the precision and reliability of baseline solution. In a complex running environment, how to maintain the long-term stability of a resolving standard, how to effectively fuse structural mechanics behaviors with real-time observation data and how to establish a closed-loop monitoring system from real-time perception to trend prediction become an important direction for further improving the intelligent level of dam safety monitoring, so that a real-time dynamic resolving system which can adaptively maintain a reference frame, fuse multi-source information and has intelligent feedback adjustment capability is developed, and the real-time dynamic resolving system has important significance for realizing the dam safety monitoring which is more reliable, more accurate and has prospective early warning capability. Disclosure of Invention The invention aims to solve the problems in the background art and provides a GNSS real-time dynamic calculation system and method for automatic monitoring of dam deformation. The technical scheme of the invention is that the GNSS real-time dynamic calculation method for automatically monitoring the deformation of the dam comprises the following concrete implementation steps: s1, dynamically sensing the running state and resolving behavior of each monitoring node in a GNSS monitoring network, automatically forming and dynamically updating resolving reference roles based on the observation continuity, resolving stability and environment response multidimensional state information of the nodes; S2, constructing a reference node time sequence evolution model, carrying out inter-node relative coordinate residual error statistics and quantification on a selected reference node set, introducing an environmental disturbance factor for correction, generating a self-correction coordinate increment to update a reference node coordinate, and carrying out overall reliability assessment on an updated reference field; S3, performing cross-epoch coupling resolving, constructing a multi-epoch state vector, performing state prediction by utilizing a reference field evolution trend, coupling real-time observation node data with self-correction reference node coordinates to construct a dynamic gain matrix, updating the node state vector, calculating a resolving residual error, and generating node resolving credibility; S4, coupling a multi-epoch GNSS (Global navigation satellite System) resolving result with dam structure mechanics constraint, constructing a structure mechanics constraint model, calculating displacement residual errors, identifying abnormal nodes by combining node credibility, carrying out weighted least square closed loop correction, dynamically updating node weights, and evaluating the whole network correction displacement consistency; And S5, based on the corrected node displacement sequence, the historical trend data and the environmental disturbance information, constructing a multi-model fusion prediction system to realize future displacement prediction, calculating a node future displacement super-threshold risk index, weighting according to node credibility to generate a whole network risk index, dividing the early warning level according to the index, and feedi