Dam aging assessment method and system based on GNSS data timing feature analysis

Abstract

The present application relates to the technical field of dam monitoring, and more particularly to a dam aging evaluation method and system based on GNSS data time sequence feature analysis, which establishes a dam monitoring database. Based on geographic environmental information, aging driving load is evaluated, hydrological load, geological disturbance and climate erosion sequence are constructed, and load influence is quantified through comprehensive sensitivity threshold. Combined with basic structure parameters, design information and GNSS time sequence features, an aging simulation space is constructed to dynamically simulate the dam aging process. The aging driving load is loaded in the simulation space, the accelerated aging time is set, and the aging evolution is simulated in stages, and the long-term aging effect is quickly reproduced in combination with the time acceleration factor. Through deep learning, the mapping relationship between GNSS time sequence features and multi-source parameters is mined, the database is updated in real time, and the model is calibrated, and evaluation results such as aging grade, key weak area and remaining safe operation life are output, thereby providing accurate decision support for dam life cycle management.

Description

Technical Field

[0001] This invention relates to the field of dam monitoring technology, and in particular to a method and system for assessing dam aging based on the temporal characteristics analysis of GNSS data. Background Technology

[0002] As a core infrastructure of water conservancy projects, the safety of dams directly affects the stable operation of important functions such as regional flood control, water supply, and power generation. With increasing service life, dam structures are subject to the coupled effects of multiple factors, including hydrological cycles, geological activity, and climate change, gradually exhibiting aging phenomena such as material deterioration, crack propagation, and foundation deformation, seriously threatening project safety. Traditional dam aging assessments mainly rely on manual inspections, single-point instrument monitoring, and static data threshold analysis, which are insufficient to comprehensively capture the dynamic evolution of structural performance under complex environments. These methods suffer from problems such as assessment lag, insufficient analysis of the coupling effects of multiple factors, and low accuracy in long-term performance prediction.

[0003] In recent years, Global Navigation Satellite Systems (GNSS) have become an important means of monitoring dam deformation due to their high-precision, all-weather, and automated displacement monitoring capabilities. The time-series parameters contained in GNSS data, such as cumulative displacement, displacement rate, periodic fluctuations, and abrupt changes, can intuitively reflect the deformation trend and abnormal response of the dam structure. However, current technologies have not fully explored the intrinsic correlation between the time-series characteristics of GNSS data and the dam aging process, and lack the ability to construct dynamic assessment models that combine multiple factors with geographic environmental information, making it difficult to achieve full life-cycle simulation and risk prediction of the dam aging process.

[0004] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0005] The main objective of this invention is to provide a method and system for dam aging assessment based on the temporal characteristics analysis of GNSS data. This aims to solve the technical problems of traditional dam aging assessment relying on manual inspection, single-point monitoring and static analysis, which suffer from assessment lag, insufficient multi-factor coupling analysis, low long-term prediction accuracy, and the fact that existing technologies do not fully explore the inherent relationship between the temporal characteristics of GNSS data and geographical environment information, lack a multi-factor coupling dynamic assessment model, and are difficult to achieve full life cycle simulation and risk prediction.

[0006] To achieve the above objectives, this invention provides a dam aging assessment method based on GNSS data time-series characteristic analysis, the method comprising:

[0007] A dam monitoring database is established, which includes basic structural parameter information of the dam, geographical environment information, dam structural design information, and historical monitoring data after the dam has been in operation. The historical monitoring data includes time-series characteristic parameters of GNSS monitoring data.

[0008] The geographic environmental information includes hydrological environmental information, geological environmental information, and climatic environmental information, and the aging-driven load is assessed based on the hydrological environmental information, geological environmental information, and climatic environmental information.

[0009] Based on the basic structural parameter information, dam structural design information, and GNSS data time series characteristic parameters, an aging simulation space is constructed for the dam to dynamically simulate the aging process of the dam.

[0010] The aging-driven load is loaded in the aging simulation space, and an accelerated aging time is set. Based on the accelerated aging time, several aging stages of the dam are dynamically simulated, and the dam aging assessment results are obtained through the aging simulation results.

[0011] Optionally, assessing the aging-driven load based on the hydrological, geological, and climatic environmental information includes:

[0012] The hydrological load stages are divided according to the hydrological environment information. The hydrological load stages include the high water level stage during the wet season, the low water level stage during the dry season, and the flood impact stage. The stages are sorted from high to low according to the load intensity to generate a hydrological load sequence.

[0013] The geological disturbance stages are divided according to the geological environment information. The geological disturbance stages include the seismic activity stage, the foundation settlement stage, and the fault sliding stage. The disturbances are then sorted from high to low intensity to generate a geological disturbance sequence.

[0014] Based on the climate environment information, the climate erosion stages are divided, including the high-temperature weathering stage, the freeze-thaw cycle stage, and the dry-wet alternation stage, and are sorted from high to low according to the erosion intensity to generate a climate erosion sequence.

[0015] The hydrological load sequence, geological disturbance sequence, and climate erosion sequence are arranged and combined in three dimensions. The results of the arrangement and combination are weighted and sorted based on the intensity of load, disturbance, and erosion to generate a comprehensive driving load sequence.

[0016] Based on the dam monitoring database, a comprehensive sensitivity threshold is set, and the comprehensive driving load sequence is evaluated for aging driving load using the comprehensive sensitivity threshold.

[0017] Optionally, the step of assessing the aging-driven load of the comprehensive driving load sequence using the comprehensive sensitivity threshold includes:

[0018] Traverse the comprehensive driving load sequence to determine whether there is a load item that matches the comprehensive sensitivity threshold, wherein the load item is a combination of single environmental effects in a three-dimensional permutation and combination;

[0019] If it exists, the comprehensive sensitivity threshold is marked on the corresponding load item, and the proportion of load items higher than the comprehensive sensitivity threshold and the proportion of load items lower than the comprehensive sensitivity threshold are determined. The aging-driven load is evaluated based on the proportion.

[0020] If none exists, then select any of the load items and determine the relationship between the combined effect intensity of the load item and the combined sensitivity threshold.

[0021] If the overall sensitivity threshold is higher, the overall driving load sequence is marked as a high-risk load sequence; if the overall sensitivity threshold is lower, the overall driving load sequence is marked as a low-risk load sequence.

[0022] Optionally, assessing the aging-driven load based on the percentage includes:

[0023] The design reference period of the dam is obtained, and the aging rate of the dam under each load item is calculated based on the comprehensive sensitivity threshold and proportion. The durability performance degradation trend of the dam structure is evaluated in conjunction with the design reference period.

[0024] A short-term impact recovery threshold is set, which is the critical value of the deformation recovery ability of the dam material under short-term high load. The dam is subjected to impact simulation using a load term higher than the comprehensive sensitivity threshold within a unit time to obtain the dam's resistance to sudden damage.

[0025] The aging-driven load is evaluated based on the durability degradation trend and resistance to sudden damage performance.

[0026] Optionally, establishing a dam monitoring database includes:

[0027] Collect historical maintenance and monitoring data of the dam, including structural design schemes of dams of different types and construction years and corresponding GNSS monitoring data. The time series characteristic parameters of the GNSS data include cumulative displacement, displacement rate, periodic fluctuation amplitude and abrupt displacement.

[0028] The historical maintenance and monitoring data information is classified according to the basic structural parameters, geographical environment information, dam structural design information, and GNSS data time series characteristic parameters. The classification includes classification by dam type, classification by construction year, and classification by geographical region. A multi-dimensional data index entry is established based on the classification results.

[0029] Structural design subsets are constructed according to each dam's structural design scheme, and a GNSS time-series feature tracking chain is established for each structural design scheme, so that the aging simulation space can call historical time-series data.

[0030] Optionally, the provision for the aging simulation space to access historical time-series data includes:

[0031] The hydrological, geological, and climatic environmental information of the dam is determined. Historical environmental parameters from the same period are matched by indexing the dam monitoring database, and environmental simulation factors are constructed. These environmental simulation factors are digital quantitative representations of the geographic environmental information in the aging simulation space.

[0032] The basic structural parameters and structural design information of the dam are determined. The structural parameters and design parameters of similar dam types are matched by indexing the dam monitoring database. The accelerated aging time is divided into stages. Each stage is set with a different time acceleration factor according to the aging rate characteristics. The time acceleration factor is positively correlated with the rate of change of GNSS time series characteristic parameters.

[0033] Based on the determined current geographical environment information and current dam structure information, the dam monitoring database is matched and the evolution law of the corresponding GNSS data time series characteristics is tracked to provide time series characteristic evolution model support for the aging simulation results.

[0034] Optionally, the dynamic simulation of several aging stages of the dam based on the accelerated aging time includes:

[0035] Set a time acceleration factor to define the conversion ratio between the length of time elapsed during the simulated aging process and the actual running time, thereby accelerating the simulation process to quickly reproduce the long-term aging effect.

[0036] Based on the actual geographical environment information, basic structural parameters and GNSS data time series characteristic parameters of the dam, several specific aging stages are defined. The aging stages include the initial micro-deformation stage, the intermediate crack development stage and the late structural damage stage, and each aging stage corresponds to a time acceleration factor.

[0037] In each defined aging stage, corresponding hydrological loads, geological disturbances, and climate erosion parameters are loaded, and the aging process is dynamically simulated by combining the dynamic changes of GNSS data time series characteristic parameters. By combining the simulation results of all the aging stages, the evolution trend of GNSS time series characteristics of the dam in the assumed whole life cycle is evaluated, and the aging simulation results are obtained.

[0038] Optionally, obtaining the dam aging assessment results through aging simulation includes:

[0039] Deep learning training is performed on the GNSS data time series feature parameters in the dam monitoring database to obtain the mapping relationship between the basic structure parameter information, geographical environment information, dam structure design information and the GNSS data time series feature parameters, as well as the nonlinear mapping relationship between the basic structure parameter information, geographical environment information, dam structure design information and the GNSS data time series feature parameters after coupling.

[0040] Set a real-time monitoring cycle, collect GNSS monitoring data of the dam in operation in real time, extract real-time time series characteristic parameters and enter them into the dam monitoring database, and dynamically update the dam monitoring database according to the real-time monitoring cycle, and adjust the mapping relationship synchronously.

[0041] Based on the mapping relationship, the GNSS time-series characteristic parameters in the aging simulation results are trend-predicted, and the prediction results are evaluated in conjunction with the dam aging level classification standard to obtain the dam aging assessment results, which include aging level, key weak areas and remaining safe operating years.

[0042] Furthermore, to achieve the above objectives, the present invention also provides a dam aging assessment system based on GNSS data time-series feature analysis. The device includes: a memory, a processor, and a dam aging assessment program based on GNSS data time-series feature analysis stored in the memory and executable on the processor. The dam aging assessment program based on GNSS data time-series feature analysis is configured to implement the steps of the dam aging assessment method based on GNSS data time-series feature analysis as described above.

[0043] In addition, to achieve the above objectives, the present invention also provides a medium storing a dam aging assessment program based on GNSS data time-series feature analysis, wherein when the dam aging assessment program based on GNSS data time-series feature analysis is executed by a processor, the dam aging assessment program based on GNSS data time-series feature analysis implements the steps of the dam aging assessment method based on GNSS data time-series feature analysis as described above.

[0044] This invention provides a dam aging assessment method based on GNSS data temporal characteristic analysis. The method constructs an aging simulation space and combines GNSS temporal characteristic parameters with dynamic environmental load application to simulate the aging process of a dam from initial micro-deformation to later structural damage in stages. Compared to traditional static threshold early warning, this method can reveal the spatiotemporal evolution of structural performance degradation in advance, such as predicting crack propagation paths or foundation settlement trends, achieving a forward-looking assessment of aging risks. By setting a time acceleration factor positively correlated with the aging rate, the aging effects of a dam over decades can be reproduced in a short time, significantly reducing the time cost required for traditional long-term monitoring. For example, by accelerating the simulation of freeze-thaw cycles or wet-dry alternations of dams in different climate zones, the cumulative damage to dams caused by extreme climates can be quickly assessed, providing real-time data support for emergency management. By using deep learning of GNSS time-series characteristics and multi-source parameters in the database to establish a nonlinear mapping model, complex coupling patterns that are difficult to detect through traditional manual analysis can be automatically captured, improving the accuracy and generalization ability of the assessment model. By combining real-time GNSS monitoring data and environmental parameters, and regularly updating the database and adjusting the mapping relationship, the assessment model can adapt to dynamic factors such as changes in material properties and environmental conditions during the dam's service life, avoiding misjudgments caused by parameter lag in traditional fixed-threshold models, and achieving continuous optimization of assessment results. Attached Figure Description

[0045] Figure 1 This is a flowchart illustrating an embodiment of the dam aging assessment method based on GNSS data temporal characteristic analysis according to the present invention.

[0046] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0047] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0048] Reference Figure 1 , Figure 1 This is a flowchart illustrating an embodiment of the dam aging assessment method based on GNSS data time-series feature analysis according to the present invention.

[0049] In one embodiment, the dam aging assessment method based on GNSS data time-series characteristic analysis includes:

[0050] Step S100: Establish a dam monitoring database. The dam monitoring database includes basic structural parameter information of the dam, geographical environment information, dam structural design information, and historical monitoring data after the dam has been in operation. The historical monitoring data includes time-series characteristic parameters of GNSS monitoring data.

[0051] The GNSS data time-series characteristic parameters can be quantitative time-series indices reflecting the change of dam displacement over time. These indices can be used as observational inputs for structural response, driving the state evolution of the aging simulation space. In this embodiment, the GNSS data time-series characteristic parameters can be obtained by continuously acquiring coordinate sequences through a GNSS receiver, and then extracting features such as cumulative displacement, displacement rate, periodic amplitude, and abrupt jumps through data filtering, trend separation, and periodic decomposition. For example, the GNSS data time-series characteristic parameters may include, but are not limited to, one or more of the following: cumulative displacement, displacement rate change rate, and seasonal periodic fluctuation amplitude.

[0052] Step S200: Geographic environmental information includes hydrological environmental information, geological environmental information and climatic environmental information, and aging-driven loads are assessed based on hydrological environmental information, geological environmental information and climatic environmental information.

[0053] The aging-driven load can be an external excitation factor acting on the deterioration process of dam materials, calculated comprehensively from hydrological, geological, and climatic environmental information. It can be used as an environmental coupling variable to trigger accelerated structural response within the aging simulation space. In this embodiment, the aging-driven load can be generated as a comprehensive load sequence through an empirical weighting model by integrating environmental data such as rainfall intensity, water level fluctuations, freeze-thaw frequency, seismic activity index, and wet-dry cycle. For example, the aging-driven load can include, but is not limited to, one or more of hydrological cycle loads, geological activity loads, and climatic cycle loads.

[0054] Step S300: Based on the basic structural parameter information, dam structural design information and GNSS data time series characteristic parameters, an aging simulation space is constructed for the dam to dynamically simulate the aging process of the dam.

[0055] The aging simulation space can be a dynamic modeling framework used to characterize the dam structure's response to environmental loads and material degradation processes in both time and space dimensions. It can be used to simulate the dam's deformation evolution process in stages and extrapolate damage paths. In this embodiment, the aging simulation space can be constructed based on basic structural parameters, structural design information, and GNSS time-series characteristic parameters to create a three-dimensional discrete mesh model, embedding a mapping relationship between time variables and state variables. Constructing the aging simulation space based on GNSS data time-series characteristic parameters can involve using GNSS time-series characteristics as input variables, along with structural parameters and design information, to input into a numerical modeling system, generating a structural state space that evolves over time. Furthermore, constructing the aging simulation space based on GNSS data time-series characteristic parameters can be achieved by using the finite element method to construct spatial discrete elements, using GNSS displacement characteristics as boundary conditions to drive element strain evolution, or by constructing a graph neural network structure, using GNSS features as node attributes and structural topology as edge relationships, and training a spatiotemporal state transfer function, thereby realizing a nonlinear mapping from observation data to the structural evolution state.

[0056] Step S400: Load the aging driving load in the aging simulation space and set the accelerated aging time. Based on the accelerated aging time, perform dynamic simulation of several aging stages of the dam and obtain the dam aging assessment results through the aging simulation results.

[0057] Loading aging-driven loads and setting accelerated aging times within the aging simulation space can be achieved by inputting the aging-driven loads as a time function into the simulation space, while simultaneously introducing a time scaling factor to compress the physical time scale and accelerate the simulation of the material degradation process. Furthermore, loading aging-driven loads and setting accelerated aging times within the aging simulation space can be achieved by employing a time recalibration algorithm to proportionally amplify the frequency of climate cycle loads and synchronously adjust the material degradation rate function, or by constructing a multi-scale time chain model to normalize the degradation rates of different environmental factors and then superimpose them onto a unified accelerated time axis. This allows for the reproduction of the structural damage accumulation process under long-term service within a finite calculation period.

[0058] Taking the freeze-thaw cycle assessment of concrete dams in high-altitude and cold regions as an example, the dam aging assessment method based on GNSS data time series characteristic analysis in this embodiment can be as follows: In a dam on a plateau, GNSS monitoring shows that the annual periodic displacement fluctuation is synchronized with the freeze-thaw cycle; the system uses this fluctuation as the GNSS time series characteristic input into the aging simulation space, and constructs a climate cycle load by combining the regional freeze-thaw frequency and water level changes. Under the action of the accelerated time factor, the ten-year freeze-thaw cycle is simulated to be completed within three weeks, and the location of crack initiation and the direction of expansion are output to guide the renewal cycle of the antifreeze coating.

[0059] This embodiment provides a dam aging assessment method based on GNSS data temporal characteristic analysis. It establishes a monitoring database containing GNSS temporal characteristic parameters, generates aging-driving loads based on environmental information, constructs an aging simulation space based on structural parameters and observation data, and loads environmental loads and acceleration time factors into the simulation space to achieve dynamic evolution simulation. By constructing the aging simulation space, the GNSS data temporal characteristic parameters and aging-driving loads are spatiotemporally coupled and mapped, allowing the dynamic evolution process of the structural response to be reconstructed into a compressible and iterative simulation sequence in the virtual space. After loading the acceleration aging time factor, the cumulative effect of environmental loads is nonlinearly amplified, and material degradation and crack propagation paths can be reproduced in short-time simulations. This process automatically mines the implicit correlation between GNSS features and environmental factors through deep learning, forming an adaptive dynamic mapping model, and continuously corrects the model under real-time data feedback. Thus, without relying on long-term observations, it achieves high-precision, low-latency prediction and assessment of the aging trend of the entire structural life cycle.

[0060] In one embodiment, the aging-driven load is assessed based on hydrological, geological, and climatic environmental information, including:

[0061] Based on hydrological environmental information, the hydrological load stages are divided into the high water level stage during the wet season, the low water level stage during the dry season, and the flood impact stage. The stages are then sorted from high to low according to the load intensity to generate a hydrological load sequence.

[0062] The hydrological load sequence can be an ordered set of stages of dam water level and water flow impact events, divided according to hydrological environmental information. It can be used to characterize the periodic and sudden effects of water pressure changes and impact forces on the dam structure. In this embodiment, the hydrological load sequence can be sorted by intensity, such as the high water level stage during the wet season, the low water level stage during the dry season, and the flood impact stage, forming an environmental action sequence with temporal priority.

[0063] Based on geological environment information, the geological disturbance stages are divided into seismic activity stages, foundation settlement stages, and fault slip stages, and are sorted from high to low disturbance intensity to generate a geological disturbance sequence.

[0064] The geological disturbance sequence can be an ordered set of abnormal activity events of the dam foundation or surrounding rock mass, classified according to geological environment information. It can be used to characterize the non-stationary disturbance of crustal movement on the stability of the dam foundation. In this embodiment, the geological disturbance sequence may include seismic activity stages, foundation settlement stages, and fault slip stages, etc., and a multi-level disturbance event sequence is constructed based on the disturbance intensity.

[0065] Based on climate and environmental information, the climate erosion stages are divided into high-temperature weathering stages, freeze-thaw cycle stages, and wet-dry alternation stages. The stages are then sorted from high to low erosion intensity to generate a climate erosion sequence.

[0066] The climate erosion sequence can be an ordered set of stages of material degradation caused by temperature and humidity changes, divided according to climate environmental information. It can be used to characterize the cumulative erosion of concrete and waterproofing materials by thermal and wet-dry cycles. In this embodiment, the climate erosion sequence can adopt stage types such as high-temperature weathering stage, freeze-thaw cycle stage, and wet-dry alternation stage, and be sorted according to erosion intensity to form an environmental degradation sequence.

[0067] The hydrological load sequence, geological disturbance sequence, and climate erosion sequence are arranged and combined in three dimensions. The results of the arrangement and combination are weighted and sorted based on the intensity of load, disturbance, and erosion to generate a comprehensive driving load sequence.

[0068] The comprehensive driving load sequence can be a multidimensional environmental load time series expression generated by three-dimensional arrangement and intensity weighting of the stage sequences of hydrological, geological, and climatic environmental factors. It can serve as a dynamic input to the aging simulation space, driving the multi-factor coupled evolution of structural degradation. In this embodiment, the comprehensive driving load sequence can be generated by arranging the hydrological load sequence, geological disturbance sequence, and climatic erosion sequence according to stage indices using a three-dimensional Cartesian product, and then weighting and sorting them according to the intensity weight of each stage to output a priority-ranked composite load event sequence. Furthermore, this process can employ a multi-objective ranking algorithm to generate a Pareto optimal combination sequence with intensity as the optimization objective, or construct an environmental factor interaction matrix and calculate and rank the comprehensive weights of each combination event using the entropy weight method. This achieves a nonlinear coupled quantitative expression of multi-source environmental effects, avoiding evaluation bias dominated by a single factor.

[0069] Based on the dam monitoring database, a comprehensive sensitivity threshold is set, and the aging driving load is evaluated based on the comprehensive driving load sequence using the comprehensive sensitivity threshold.

[0070] In this embodiment, the comprehensive driving load sequence can be used as a time series input, driving the state transition function in the aging simulation space according to event priority and timestamp. Each load event is mapped to an incremental function of material degradation rate, which is superimposed on the state equation of the aging model. Alternatively, an event-driven simulation mechanism can be used to update the simulation state only when a load event is triggered, reducing computational redundancy and enabling the simulation process to respond to the temporal logic of coupled events in the real environment, thereby improving the physical consistency of the damage evolution path. Taking the composite damage simulation of a concrete dam in an earthquake-prone area as an example, this embodiment could be a dam area that has experienced a strong earthquake and then entered a period of continuous rainfall. The system combines the seismic activity stage and the high water level stage of the high water period into a high-weight composite event, superimposed with the freeze-thaw cycle stage to form a triple coupled load. This comprehensive driving load sequence is loaded into the aging simulation space, triggering the accelerated initiation of microcracks in the concrete impermeable layer under the combined action of high water pressure and frost heave stress. The model outputs the crack propagation path within 24 hours, indicating that leakage detection should be carried out immediately.

[0071] This embodiment divides the phase sequences of three types of environmental factors—hydrology, geology, and climate—and sorts them by intensity. The three sequences are then arranged in three dimensions and weighted by intensity to generate a comprehensive driving load sequence. Aging assessment is then performed based on a comprehensive sensitivity threshold, transforming the multi-source environmental coupling effect from isolated analysis to a systematic temporal expression. This sequence, as a high-dimensional input, is loaded into the aging simulation space, forming a dynamic coupling with GNSS temporal characteristics. This drives the model to accurately reproduce real composite damage scenarios such as "earthquake + high water level + freeze-thaw." Combined with a time acceleration factor, this mechanism restores long-term cumulative effects under compressed time scales, while providing structured and traceable environmental inputs for deep learning. This enables the assessment model to generalize to cope with complex alternating environmental effects, achieving a technological upgrade from single-factor response prediction to multi-field coupled evolution prediction.

[0072] In one embodiment, aging-driven load assessment of the comprehensive driving load sequence is performed using a comprehensive sensitivity threshold, including: traversing the comprehensive driving load sequence and determining whether there is a load item that matches the comprehensive sensitivity threshold, wherein the load item is a combination of single environmental effects in a three-dimensional permutation.

[0073] The comprehensive sensitivity threshold can be a dynamic quantitative standard used to determine the risk level of environmentally coupled events in a comprehensive driving load sequence. It can serve as a basis for risk classification, distinguishing between high-risk and low-risk combinations of environmental effects. In this embodiment, the comprehensive sensitivity threshold can be initialized based on historical monitoring data and expert experience, and dynamically calibrated by statistically analyzing the intensity distribution of high-risk events in the training samples. For example, the comprehensive sensitivity threshold can include, but is not limited to, one or more of structural response sensitivity thresholds, material degradation sensitivity thresholds, and environmental synergy sensitivity thresholds. The comprehensive driving load sequence can be a multidimensional environmental load time series expression generated by three-dimensional arrangement and intensity weighting of hydrological, geological, and climatic environmental factor stage sequences. It can be used as a dynamic input to the aging simulation space, driving the multi-factor coupled evolution of structural degradation. In this embodiment, the comprehensive driving load sequence can be arranged by Cartesian product of the three stage sequences, weighted according to the intensity of each stage, and then globally sorted and output. For example, the comprehensive driving load sequence can include, but is not limited to, one or more of hydrological-dominated composite loads, geological-dominated composite loads, and climatic-dominated composite loads.

[0074] Traversing the comprehensive driving load sequence and determining the risk level of load items based on the comprehensive sensitivity threshold can be achieved by examining each three-dimensional environmental combination in the comprehensive driving load sequence item by item, comparing the relative magnitude of its comprehensive intensity and the comprehensive sensitivity threshold, and marking it as high-risk or low-risk. Furthermore, traversing the comprehensive driving load sequence and determining the risk level of load items based on the comprehensive sensitivity threshold can be achieved by using a sliding window mechanism to perform local risk clustering analysis on consecutive load items in the sequence. This allows for fine-grained risk identification of complex coupled events and avoids misjudgment based on a single indicator.

[0075] If it exists, the comprehensive sensitivity threshold is marked on the corresponding load item, and the proportion of load items above the comprehensive sensitivity threshold and the proportion of load items below the comprehensive sensitivity threshold are determined. The aging-driven load is evaluated based on the proportion.

[0076] Assessing aging-driven load based on the proportion of load items above / below a comprehensive sensitivity threshold can be achieved by calculating the proportion of load items above the threshold in the comprehensive driving load sequence, serving as a quantitative indicator of overall aging-driven intensity. Furthermore, assessing aging-driven load based on the proportion of load items above / below the comprehensive sensitivity threshold can be achieved by fitting the risk proportion using a cumulative distribution function and outputting a risk confidence interval. This allows high-dimensional environmental coupling information to be compressed into an interpretable single-dimensional risk indicator, supporting the stability assessment of model inputs.

[0077] If none exists, then any load item is selected, and the relationship between the combined effect intensity of the load item and the combined sensitivity threshold is determined. If it is higher than the combined sensitivity threshold, the combined driving load sequence is marked as a high-risk load sequence; if it is lower than the combined sensitivity threshold, the combined driving load sequence is marked as a low-risk load sequence. Taking the coupled assessment of wet-dry alternation and foundation settlement of gravity dams in arid areas as an example, the dam aging assessment method based on GNSS data time series feature analysis in this embodiment can be that the combined driving load sequence contains 128 environmental combinations, of which only 3 combinations exceed the combined sensitivity threshold, accounting for 2.3%. The system determines that the overall risk is low, but identifies the combination of 'wet-dry alternation + foundation settlement' as a high-risk core mode. This mode is marked and input into the aging simulation space, triggering the microcrack evolution simulation of the concrete tensile stress zone, and providing an early warning of the dam tilting trend caused by non-uniform settlement 11 months in advance.

[0078] This embodiment traverses the comprehensive driving load sequence and determines the risk level of load items based on the comprehensive sensitivity threshold. It assesses the aging driving load based on the proportion of load items above or below the comprehensive sensitivity threshold, and determines the overall risk level through single-point intensity comparison when there are no matching items. This achieves intelligent hierarchical assessment of multi-dimensional environmental coupling events by introducing a comprehensive sensitivity threshold. It is the first to realize the quantitative analysis of complex coupling risk distribution, identify the key modes of dominant degradation, and transform nonlinear environmental coupling into a statistically significant and traceable risk distribution model. It provides dynamic input with physical meaning and statistical significance for the aging simulation space, enabling the assessment system to identify global evolution trends from local anomalies, supporting the forward-looking prediction of crack propagation and settlement trends, and achieving the technical upgrade from passive threshold triggering to active risk diagnosis.

[0079] In one embodiment, assessing the aging-driven load based on percentage includes:

[0080] Obtain the design reference period of the dam, calculate the aging rate of the dam under each load item based on the comprehensive sensitivity threshold and proportion, and evaluate the degradation trend of the durability performance of the dam structure in conjunction with the design reference period.

[0081] The comprehensive sensitivity threshold can be a dynamic quantitative standard used to determine the risk level of environmentally coupled events in a comprehensive driving load sequence. It can also be used as a trigger condition for calculating the aging rate, linking environmental load intensity with structural degradation rate. In this embodiment, the comprehensive sensitivity threshold can be initially set based on historical monitoring data and expert experience, and dynamically calibrated by statistically analyzing the intensity distribution of high-risk events in the training samples. For example, the comprehensive sensitivity threshold can include, but is not limited to, one or more of structural response sensitivity thresholds, material degradation sensitivity thresholds, and environmental synergy sensitivity thresholds. Calculating the aging rate of the dam under each load item based on the comprehensive sensitivity threshold and its proportion, and evaluating the durability degradation trend in conjunction with the design reference period, can involve assigning aging weights to load items in the comprehensive driving load sequence that exceed the comprehensive sensitivity threshold based on their proportion, and calculating the cumulative damage proportion in conjunction with the design reference period. Furthermore, this operation can be achieved by using a time-cumulative damage model to map the product of load item intensity and proportion to an annual average degradation rate, which is then accumulated to the design reference period. This allows for quantitative evolution modeling of the impact of environmentally coupled loads on long-term structural durability.

[0082] A short-term impact recovery threshold is set, which is the critical value of the deformation recovery capacity of the dam material under short-term high load. The dam is subjected to impact simulation using a load term higher than the comprehensive sensitivity threshold within a unit time to obtain the dam's resistance to sudden damage.

[0083] The short-term impact recovery threshold can be a critical value characterizing the ability of dam materials to recover their original deformation after a short period of high load. It can be used as a criterion for judging the performance against sudden damage, reflecting the elastic recovery potential of the material under impact load. In this embodiment, the short-term impact recovery threshold can be obtained by using laboratory accelerated loading tests to obtain the stress-strain recovery curve of the material under transient load and extracting the critical load strength when the recovery rate is lower than a set value. For example, the short-term impact recovery threshold can include, but is not limited to, one or more of the following: elastic recovery critical value, plastic residual critical value, microcrack closure critical value, etc. Using load terms higher than the comprehensive sensitivity threshold to simulate the impact on the dam to obtain the performance against sudden damage can be done by selecting a single or combined load term higher than the comprehensive sensitivity threshold in the comprehensive driving load sequence within the unit time scale, applying it to the aging simulation space, and observing whether the material deformation recovery is lower than the short-term impact recovery threshold. Furthermore, this operation can be achieved by applying transient load pulses in the finite element model and recording whether the residual strain exceeds the threshold after unloading, thereby obtaining the toughness response characteristics of the material under extreme coupled impact and distinguishing between slow degradation and sudden failure modes.

[0084] Taking the toughness assessment of a high-altitude reservoir dam under flood impact as an example, the dam aging assessment method based on GNSS data time series characteristic analysis in this embodiment can be that the combined proportion of 'flood impact + freeze-thaw cycle' in the comprehensive driving load sequence reaches 8.7%, triggering accelerated calculation of aging rate; the system applies this combined impact in the simulation unit, and observes that the concrete strain recovery rate drops to 12%, which is lower than the short-term impact recovery threshold of 15%; it is determined that this coupled mode has the risk of sudden damage, and the damage evolution equation of the aging simulation space is updated synchronously to provide early warning of irreversible cracking that may occur in the dam's anti-seepage layer before the next flood peak.

[0085] This embodiment evaluates the durability performance degradation trend by obtaining the design baseline period and calculating the aging rate based on the comprehensive sensitivity threshold and load ratio. By setting a short-term impact recovery threshold and applying high-risk load impacts to obtain the performance against sudden damage, it can quantify the damage intensity of environmentally coupled loads through the comprehensive sensitivity threshold and define the material's resistance to sudden damage by combining the short-term impact recovery threshold. This is the first time that a two-dimensional evaluation mechanism for durability performance degradation trend and impact toughness performance has been constructed. On the one hand, the long-term cumulative degradation rate is calculated based on the load ratio and design baseline period to achieve quantitative prediction of the slow degradation process. On the other hand, by actively applying high-risk load impacts and comparing them with the recovery threshold, the critical failure mode of the material under extreme coupling is identified. This mechanism forms a closed-loop mapping between GNSS time-series response, environmental load sequence and material intrinsic properties, so that the aging simulation space can not only reproduce progressive damage, but also simulate the triggering conditions of sudden failure. It supports high-fidelity prediction of crack propagation path and settlement trend throughout the entire life cycle, realizing an upgrade of the evaluation paradigm from passive observation to active testing.

[0086] In some embodiments, a dam monitoring database is established, including: collecting historical maintenance and monitoring data of the dam, including structural design schemes of dams of different types and construction years and corresponding GNSS monitoring data, and the time series characteristic parameters of the GNSS data including cumulative displacement, displacement rate, periodic fluctuation amplitude and abrupt displacement.

[0087] The GNSS time-series feature tracing chain can be a structured evolutionary data sequence formed by binding structural design schemes under specific dam types, construction years, and geographical regions with their corresponding GNSS time-series feature parameters. This can be used to provide physically traceable historical response benchmark data for aging simulation space. In this embodiment, the GNSS time-series feature tracing chain can be based on the unique identifier of the structural design scheme, associating its historical GNSS cumulative displacement, displacement rate, periodic fluctuation amplitude, and abrupt displacement, etc., and concatenating them into an ordered data chain according to timestamps. For example, the GNSS time-series feature tracing chain can include, but is not limited to, one or more of the following: concrete gravity dam time-series chain, arch dam segmented time-series chain, and earth-rock dam zonal time-series chain.

[0088] The historical maintenance and monitoring data information is classified according to the basic structural parameters, geographical environment information, dam structural design information, and GNSS data time series characteristic parameters. The classification includes classification by dam type, classification by construction year, and classification by geographical region. A multi-dimensional data index entry is established based on the classification results.

[0089] The multidimensional data index entry point can be a structured data retrieval interface built based on three dimensions: dam type, construction year, and geographical region. This interface can be used to achieve semantic-based rapid location and conditional filtering of historical monitoring data. In this embodiment, the multidimensional data index entry point can establish three-dimensional index key-value pairs, mapping dam type classification codes, construction year interval codes, and geographical region codes respectively, supporting combined queries to invoke the corresponding tracking chain. For example, the multidimensional data index entry point can include, but is not limited to, one or more of the following: dam type-year-region joint index, dam type-region secondary index, and year-region secondary index.

[0090] Structural design subsets are constructed according to each dam's structural design scheme, and a GNSS time-series feature tracking chain is established for each structural design scheme, so that the aging simulation space can call historical time-series data.

[0091] The aging simulation space can be a dynamic modeling framework used to characterize the dam structure's response to environmental loads and material degradation processes in both time and space dimensions. It can be used to simulate the dam's deformation evolution process in stages and extrapolate damage paths. In this embodiment, the aging simulation space can construct a three-dimensional discrete mesh model based on basic structural parameters, structural design information, and GNSS time-series characteristic parameters, embedding a mapping relationship between time variables and state variables. For example, the aging simulation space can include, but is not limited to, one or more of the following: linear elastic aging model, nonlinear viscoelastic-plastic model, and multi-field coupled damage evolution model.

[0092] Establishing a GNSS timing feature tracing chain for each structural design scheme can be achieved by binding a specific structural design scheme to its historical GNSS timing feature parameters in chronological order, forming a structured data sequence that can be independently accessed. Furthermore, this GNSS timing feature tracing chain can be stored using a graph database, with the structural design ID as the root node and timing features as timestamped attribute edges, or by constructing a JSON serialized file set, where each file corresponds to a structural scheme and includes timing parameters and metadata tags. This allows for precise coupling between structural design and observation response, supporting the physical consistency of simulation inputs.

[0093] Accessing historical time-series data for aging simulation space via a multi-dimensional data index can be achieved by querying based on a combination of three conditions: dam type, construction year, and geographical region. This allows for the location and extraction of matching GNSS time-series tracking chains as simulation input. Furthermore, accessing historical time-series data for aging simulation space via a multi-field SQL query can extract matching tracking chain data from a relational database, or an inverted index engine can be used to quickly retrieve and load the corresponding time-series dataset using classification codes as keys. This ensures that the input data for aging simulation space exhibits high similarity to the service environment and structural type, enhancing the representativeness of the simulation benchmark.

[0094] Taking the prediction of the aging trend of concrete gravity dams in the 1970s as an example, the dam aging assessment method based on GNSS data time series feature analysis in this embodiment can be that the system enters the multi-dimensional data index entry, inputs dam type = concrete gravity dam, construction year = 1970-1979, geographical region = North China, and retrieves three matching GNSS time series feature tracking chains; the aging simulation spatially calls the cumulative displacement and periodic fluctuation data in these chains as the initial state, combines the current climate load, and infers the crack propagation path under the accelerated time factor, outputting a damage evolution sequence that highly matches the historical maintenance records.

[0095] This embodiment collects historical maintenance and monitoring data of the dam and extracts GNSS time-series feature parameters. It establishes a binding relationship between the structural design scheme and GNSS time-series features to form a tracking chain, constructs a multi-dimensional index to achieve semantic data retrieval, and calls matching tracking chains based on classification conditions for use in the aging simulation space. This achieves a strong correlation between the structural design scheme and historical observation responses by establishing a GNSS time-series feature tracking chain, so that the input to the aging simulation space no longer relies on general statistical models but is based on evolution paths under real service conditions. The multi-dimensional data index entry ensures that the simulation input accurately matches historical data of similar structures and environmental backgrounds, giving accelerated aging simulation physical traceability and scenario reproducibility. This mechanism transforms the originally mixed monitoring data into structured, semantic evolutionary knowledge units, supporting deep learning models to mine nonlinear coupling laws in a low-noise, high-consistency data space. Thus, without relying on long-term field measurements, it achieves high-fidelity, interpretable prediction of structural damage evolution paths.

[0096] In some embodiments, the aging simulation space accesses historical time-series data, including: determining the current hydrological environment information, geological environment information and climate environment information of the dam, indexing and matching historical environmental parameters of the same period through the dam monitoring database, and constructing environmental simulation factors, which are digital quantitative representations of geographical environment information in the aging simulation space.

[0097] The environmental simulation factor can be a numerically quantified parameter that transforms hydrological, geological, and climatic environmental information into an input to the aging simulation space. It can be used to achieve a calculable expression and dynamic loading of real environmental loads within the simulation framework. In this embodiment, the environmental simulation factor can be mapped into a multi-dimensional vector based on the current hydrological fluctuation intensity, geological activity frequency, and climate cycle characteristics of the dam through a standardized weighting function, serving as an external excitation input to the simulation space. For example, the environmental simulation factor can include, but is not limited to, one or more of the following: hydrological load vector, geological disturbance index, and climate cycle intensity coefficient. Constructing the environmental simulation factor as a numerically quantified representation of geographic environmental information can be achieved by transforming hydrological, geological, and climatic environmental parameters into numerical vectors through a standardized function, which are then input into the aging simulation space as environmental excitations. Furthermore, the construction of the environmental simulation factor can be achieved by using principal component analysis to reduce the dimensionality of multi-source environmental variables and extract the dominant environmental factor composition vector, thereby enabling a calculable expression of unstructured environmental information and supporting the environmental coupling drive of the simulation space.

[0098] The basic structural parameters and design information of the dam are determined. The structural parameters and design parameters of similar dam types are matched by indexing the dam monitoring database. The accelerated aging time is divided into stages. Each stage is set with a different time acceleration factor according to the aging rate characteristics. The time acceleration factor is positively correlated with the rate of change of GNSS time series characteristic parameters.

[0099] The time acceleration factor can be a dynamic scaling factor used to compress the physical time scale and adjust the aging simulation process rate. It can be used to synchronize the simulated time series with the actual structural degradation rate, improving the physical consistency of the stage evolution logic. In this embodiment, the time acceleration factor can calculate the acceleration weight in real time based on the rate of change of GNSS time series characteristic parameters, such as displacement rate increment and mutation frequency, forming a nonlinear time scaling value positively correlated with the degradation intensity. For example, the time acceleration factor can include, but is not limited to, one or more of the following: acceleration coefficient for the micro-deformation stage, acceleration coefficient for the crack initiation stage, and acceleration coefficient for the structural damage stage. Setting a time acceleration factor positively correlated with the rate of change of GNSS time series characteristics can involve calculating the rate of change of the current GNSS characteristic parameters in real time and mapping it to a time acceleration coefficient. Furthermore, setting a time acceleration factor positively correlated with the rate of change of GNSS time series characteristics can be achieved by using a sliding window to statistically analyze the standard deviation of displacement rates as an input variable, constructing a piecewise linear function that binds the frequency of mutation events and the acceleration factor to a monotonically increasing relationship. This allows the acceleration process to adapt to the structural degradation intensity, avoiding stage distortion caused by a fixed acceleration ratio.

[0100] Based on the determined current geographic environment information and current dam structure information, the dam monitoring database is matched and the evolution law of the corresponding GNSS data time series characteristics is tracked to provide time series characteristic evolution model support for aging simulation results.

[0101] The evolution patterns of GNSS data time-series characteristics can be behavioral patterns reflecting the evolution of structural response over time, extracted from historical GNSS monitoring data of similar dams. These patterns can provide a physically based temporal evolution benchmark for the current simulation, constraining the rationality of the simulation path. In this embodiment, the evolution patterns of GNSS data time-series characteristics can be extracted by matching historical tracking chains of similar dam types, ages, and regions using a multi-dimensional index, revealing the morphological characteristics and statistical distribution of cumulative displacement, periodic fluctuations, and abrupt changes. For example, the evolution patterns of GNSS data time-series characteristics may include, but are not limited to, one or more of the following: periodic fluctuation evolution patterns, abrupt displacement clustering patterns, and nonlinear growth patterns of cumulative displacement. Tracking and matching the evolution patterns of GNSS data time-series characteristics to support aging simulation can be achieved by retrieving historical GNSS tracking chains of similar dams through a multi-dimensional index entry point, extracting their evolutionary forms as initial constraints and path references for the simulation. Furthermore, tracking and matching the evolution patterns of GNSS data temporal features to support aging simulation can be achieved by using a dynamic time warping algorithm to align the morphological similarity between historical and current temporal features, calculating the semantic distance between the historical tracking chain and the current state based on a graph embedding model, and selecting the most similar trajectory. This ensures that the simulated evolution path is based on the real historical response and improves the physical interpretability of the prediction.

[0102] Taking the simulation of alternating wet and dry damage of earth-rock dams in the hot and humid southern region as an example, the dam aging assessment method based on GNSS data time series characteristic analysis in this embodiment can be as follows: the system detects the intensity of the current water level fluctuation in the dam body and matches the rainfall cycle with historical data of the same period to generate a hydrological-climate composite environment simulation factor; based on the sudden increase in GNSS displacement rate, the time acceleration factor is automatically increased to 3.2 times; the GNSS mutation clustering pattern of similar earth-rock dams in similar climate zones is retrieved as the evolution benchmark, and the chain evolution of surface seepage-cracking-settlement is reproduced in the simulation space to output the potential slip risk area for the next 6 months.

[0103] This embodiment constructs an environmental simulation factor as a digital quantitative representation of geographic environmental information. By setting a time acceleration factor positively correlated with the rate of change of GNSS temporal characteristics, and by tracking and matching the evolution law of GNSS data temporal characteristics to support aging simulation, it can achieve the digital quantitative injection of real environmental loads through the environmental simulation factor, dynamically adjust the aging process according to the GNSS change rate through the time acceleration factor, and provide physical constraint paths by tracking historical temporal evolution laws. The three elements work together to construct a three-dimensional simulation mechanism of environment-driven, rate-adaptive, and historical trajectory-constrained simulation. This mechanism makes the simulation space no longer dependent on static models or fixed speedup ratios, but dynamically extrapolates based on the current structural response intensity and historical similar evolution patterns, significantly improving the physical consistency of aging stage transitions and the interpretability of predicted paths. Thus, without relying on long-term field measurements, it achieves the technical effect of high-fidelity, adaptive, and traceable full life cycle aging assessment.

[0104] In one embodiment, dynamic simulations are performed on several aging stages of the dam based on accelerated aging time, including:

[0105] Set a time acceleration factor to define the conversion ratio between the length of time elapsed during the simulated aging process and the actual running time, thereby accelerating the simulation process to quickly reproduce the long-term aging effect.

[0106] Based on the actual geographical environment information, basic structural parameters and GNSS data time series characteristic parameters of the dam, several specific aging stages are defined. The aging stages include the initial micro-deformation stage, the intermediate crack development stage and the later structural damage stage, and each aging stage corresponds to a time acceleration factor.

[0107] At each defined aging stage, corresponding hydrological loads, geological disturbances, and climate erosion parameters are loaded, and the aging process is dynamically simulated by combining the dynamic changes of GNSS data time series characteristic parameters. By combining the simulation results of all aging stages, the evolution trend of GNSS time series characteristics of the dam in the assumed whole life cycle is evaluated, and the aging simulation results are obtained.

[0108] The time acceleration factor can be a scaling factor used to quantify the conversion ratio between simulated time and actual service time. It can be used to compress the physical time scale within a specific aging stage, enabling rapid simulation of long-term degradation effects. In this embodiment, the time acceleration factor can be derived as a stage-specific scaling factor based on the material degradation rate and environmental load intensity at each stage through empirical regression or machine learning fitting. For example, the time acceleration factor can include, but is not limited to, one or more of the following: acceleration factor for the micro-deformation stage, acceleration factor for the crack development stage, and acceleration factor for the structural damage stage.

[0109] The aging stage can be a phased state interval with a clear physical form and dominant mechanism, defined based on the characteristics of dam structural deterioration and evolution. It can be used to deconstruct the continuous aging process into discrete evolutionary units that can be independently modeled and parameterized. In this embodiment, the aging stage can be divided according to the morphological abrupt change points of GNSS time-series characteristics, the cumulative threshold of environmental loads, and the nonlinear transition characteristics of the structural response. For example, the aging stage can include, but is not limited to, one or more of the following: the initial micro-deformation stage, the mid-term crack development stage, and the late-stage structural damage stage. GNSS data time-series characteristic parameters can be quantitative time-series indicators obtained by the GNSS monitoring system, reflecting the change of dam displacement over time. These can be used as dynamic input signals for the structural response within each aging stage, driving the state transition of the evolution model within that stage. In this embodiment, GNSS data time-series characteristic parameters can include specific quantitative indicators such as cumulative displacement, displacement rate change rate, and seasonal periodic fluctuation amplitude.

[0110] Based on geographic environmental information, infrastructure parameters, and temporal characteristic parameters of GNSS data, several aging stages can be defined. This can be achieved by combining multi-source environmental data and GNSS response sequences to identify significant segmentation inflection points in structural behavior and delineate stage boundaries with independent degradation mechanisms. Furthermore, defining these aging stages based on geographic environmental information, infrastructure parameters, and temporal characteristic parameters of GNSS data can be achieved by using change point detection algorithms to perform piecewise regression on GNSS displacement sequences, minimizing residuals to determine stage segmentation points, or by using cluster analysis to perform unsupervised classification of the joint environmental-response feature space, generating stage prototypes and membership relationships. This allows for a structural transformation of the aging process from continuous observation to discrete mechanism modeling.

[0111] Dynamic simulations can be performed by loading corresponding hydrological loads, geological disturbances, and climate erosion parameters at each aging stage and combining them with temporal characteristic parameters of GNSS data. This can be achieved by assigning a dedicated environmental load sequence and time acceleration factor to each aging stage, using GNSS features as response variables input into the evolution model, and performing state extrapolation within each stage. Furthermore, dynamic simulations can be performed by loading corresponding hydrological loads, geological disturbances, and climate erosion parameters at each aging stage and combining them with temporal characteristic parameters of GNSS data. This can be achieved by switching material constitutive relations and boundary conditions in the finite element model according to each stage, synchronously updating the load frequency and amplitude, or by constructing a stage-aware LSTM network, using environmental parameters as input and GNSS features as output, and training a stage-specific temporal mapping function. This allows for the accurate reproduction of multi-factor coupled degradation mechanisms under staged conditions.

[0112] Taking the multi-stage aging assessment of concrete dams as an example, the dam aging assessment method based on the temporal characteristic analysis of GNSS data in this embodiment can be as follows: GNSS data of a concrete dam shows that the displacement increases linearly in the first 10 years, accelerates periodically from 10 to 25 years, and then frequently changes abruptly after 25 years. Based on this, the system divides the dam into three stages: initial, middle and late, and assigns time acceleration factors of 1:5, 1:15 and 1:30 respectively. The initial stage is loaded with water level fluctuations and temperature gradients, the middle stage is superimposed with freeze-thaw cycles and chemical erosion, and the late stage is introduced with seismic disturbances and seepage pressure. The model completes the full life simulation within 6 months and outputs the path evolution map of cracks from the surface to the deep layer.

[0113] This embodiment achieves rapid simulation of long-term degradation effects by setting a time acceleration factor to compress the physical time scale. It divides the dam into phased state intervals with independent degradation mechanisms based on geographical environmental information, basic structural parameters, and GNSS data time-series characteristics. By assigning specific environmental loads and time acceleration factors to each phase and dynamically extrapolating using GNSS time-series characteristics, it achieves segmented and accurate modeling of the dam's entire life-cycle degradation process by defining physically interpretable aging stages and configuring specific parameter combinations. In each phase, GNSS time-series characteristics serve as dynamic response signals driving the nonlinear evolution of multi-source environmental factors, ensuring that the simulation process retains both measured data and accelerated reproducibility. Phase boundaries and acceleration factors can be dynamically redefined with real-time data, forming a closed-loop feedback mechanism. This achieves the technical effect of phased, interpretable, and highly timely prediction of structural damage paths without relying on long-term observations.

[0114] In some embodiments, aging assessment results of dams are obtained through aging simulation results, including: performing deep learning training on the time-series characteristic parameters of GNSS data in the dam monitoring database to obtain the mapping relationships between the basic structural parameter information, geographical environment information, and dam structural design information and the time-series characteristic parameters of GNSS data, as well as the nonlinear mapping relationship between the coupled basic structural parameter information, geographical environment information, and dam structural design information and the time-series characteristic parameters of GNSS data.

[0115] The nonlinear mapping relationship can be a function expression learned through a deep learning model, describing the complex nonlinear relationship between multi-source input parameters and GNSS temporal feature parameters. It can be used to automate the modeling and predictive simulation of the coupling mechanism between environment, structure, and response. In this embodiment, the nonlinear mapping relationship can adopt a multi-layer neural network architecture, using basic structural parameter information, geographic environment information, and dam structural design information as input layers, and GNSS data temporal feature parameters as output layers. A high-dimensional mapping function is established by optimizing the weight parameters through backpropagation. For example, the nonlinear mapping relationship can include, but is not limited to, one or more of the following: single-factor independent mapping relationship, two-factor interactive mapping relationship, and fully parameter-coupled mapping relationship.

[0116] Set a real-time monitoring cycle, collect GNSS monitoring data of the dam in operation in real time, extract real-time time series characteristic parameters and enter them into the dam monitoring database, and dynamically update the dam monitoring database according to the real-time monitoring cycle, and adjust the mapping relationship synchronously.

[0117] The dam monitoring database can be a dynamic knowledge base that centrally stores dam structural parameters, environmental information, and GNSS time-series observation data. It can serve as a data carrier for training and updating mapping relationships, supporting the online adaptive capabilities of the model. In one specific embodiment, the dam monitoring database can continuously access raw GNSS coordinate sequences, hydrological station monitoring values, meteorological station records, and structural design documents through automated data interfaces, and store them in a unified format according to timestamps. Furthermore, the dam monitoring database can include, but is not limited to, a static basic information database, historical monitoring datasets, and real-time streaming data caches.

[0118] Based on the mapping relationship, the GNSS time-series characteristic parameters in the aging simulation results are trend-predicted, and the prediction results are evaluated in conjunction with the dam aging level classification standard to obtain the dam aging assessment results, which include aging level, key weak areas and remaining safe operating years.

[0119] In this embodiment, the GNSS data time-series characteristic parameters can be quantitative time-series indicators that reflect the change of dam displacement over time, obtained by the GNSS monitoring system. These indicators can be used as output targets and update trigger signals for mapping relationships, driving the model to perceive the structural state in real time. For example, the GNSS data time-series characteristic parameters may include cumulative displacement, displacement rate of change, and seasonal periodic fluctuation amplitude.

[0120] Deep learning is used to train the temporal feature parameters of GNSS data in the dam monitoring database to obtain the nonlinear mapping relationship between multi-source parameters and GNSS temporal features. This can be achieved by training a neural network model using historical datasets, enabling the model to learn the nonlinear correspondence between input parameter combinations and GNSS feature outputs. Furthermore, this operation can be achieved by using a Transformer architecture to encode multi-source heterogeneous parameter sequences and focusing on key coupling periods with an attention mechanism, or by constructing a graph convolutional network with the dam structural topology as graph nodes and the environment and GNSS features as node attributes, learning a joint space-time mapping. This transforms traditional manual experience-based judgments into a generalizable, data-driven prediction function.

[0121] The dam monitoring database is dynamically updated based on real-time monitoring cycles, and the nonlinear mapping relationship is adjusted synchronously. This can be achieved by collecting new GNSS data and environmental parameters at preset intervals, injecting them into the database, triggering incremental model training, and fine-tuning the mapping relationship weights to adapt to the latest state. Furthermore, this operation can be implemented by using online learning algorithms to update only the parameter gradients corresponding to the most recent window of data without retraining the entire model, or by constructing a dual-channel model architecture where the main model remains stable while the secondary model is fine-tuned in real time. The final mapping result is then output through weighted fusion. This allows for the evaluation of the model's real-time response and adaptive correction to dynamic factors such as material degradation and abrupt climate changes.

[0122] Taking the dynamic assessment of a 20-year-old concrete dam as an example, the dam aging assessment method based on GNSS data time series feature analysis in this embodiment can be that a dam monitoring database has accumulated 15 years of GNSS data, and the initial mapping relationship is trained based on historical data; every 7 days, new GNSS displacement sequences and rainfall and reservoir water level data are automatically collected and injected into the database to trigger incremental training of a lightweight model; the new mapping relationship captures the abnormal upward trend of displacement rate in the seepage zone, and the model automatically outputs a warning of "accelerated development of mid-term cracks". Combined with the aging level standard, it is judged as a level III risk, predicting that the remaining safe operating years are 8-12 years, and locating the downstream dam abutment as a key weak area.

[0123] This embodiment establishes a nonlinear mapping relationship between multi-source parameters and GNSS time-series characteristics through deep learning training. The mapping relationship is continuously optimized through real-time acquisition and dynamic updating mechanisms. The evaluation results are output by combining trend prediction with grading standards. This achieves high-dimensional coupled modeling of the multi-source structure and environmental parameters' response to GNSS time-series responses by constructing a deep learning-based nonlinear mapping relationship. Combined with the dynamic updating mechanism of the dam monitoring database, the mapping relationship continuously absorbs real-time observation data, forming a closed-loop adaptive system. Based on aging simulation results, this system predicts GNSS evolution trends using the latest mapping relationship and outputs aging levels, weak areas, and remaining lifespan according to grading standards. This transforms the evaluation from static deduction to real-time diagnosis based on actual dynamic responses, overcoming the lag and misjudgment caused by parameter fixation in traditional models. This achieves the technical effect of high-precision, adaptive, and interpretable full life-cycle evaluation capabilities.

[0124] Furthermore, to achieve the above objectives, the present invention also provides a dam aging assessment system based on GNSS data time-series feature analysis. The device includes: a memory, a processor, and a dam aging assessment program based on GNSS data time-series feature analysis stored in the memory and executable on the processor. The dam aging assessment program based on GNSS data time-series feature analysis is configured to implement the steps of the dam aging assessment method based on GNSS data time-series feature analysis as described above.

[0125] In addition, to achieve the above objectives, the present invention also provides a medium storing a dam aging assessment program based on GNSS data time-series feature analysis, wherein when the dam aging assessment program based on GNSS data time-series feature analysis is executed by a processor, the dam aging assessment program based on GNSS data time-series feature analysis implements the steps of the dam aging assessment method based on GNSS data time-series feature analysis as described above.

[0126] Other embodiments or specific implementations of the dam aging assessment system based on GNSS data time-series characteristic analysis described in this invention can be referred to the above-mentioned method embodiments, and will not be repeated here.

[0127] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A method for assessing dam aging based on GNSS data time-series characteristic analysis, characterized in that, The method includes: A dam monitoring database is established, which includes basic structural parameter information of the dam, geographical environment information, dam structural design information, and historical monitoring data after the dam has been in operation. The historical monitoring data includes time-series characteristic parameters of GNSS monitoring data. The geographic environmental information includes hydrological environmental information, geological environmental information, and climatic environmental information, and the aging-driven load is assessed based on the hydrological environmental information, geological environmental information, and climatic environmental information. Based on the basic structural parameter information, dam structural design information, and GNSS data time series characteristic parameters, an aging simulation space is constructed for the dam to dynamically simulate the aging process of the dam. The aging-driven load is loaded in the aging simulation space, and an accelerated aging time is set. Based on the accelerated aging time, several aging stages of the dam are dynamically simulated, and the dam aging assessment results are obtained through the aging simulation results.

2. The dam aging assessment method based on GNSS data time-series characteristic analysis as described in claim 1, characterized in that, The assessment of aging-driven loads based on the hydrological, geological, and climatic environmental information includes: The hydrological load stages are divided according to the hydrological environment information. The hydrological load stages include the high water level stage during the wet season, the low water level stage during the dry season, and the flood impact stage. The stages are sorted from high to low according to the load intensity to generate a hydrological load sequence. The geological disturbance stages are divided according to the geological environment information. The geological disturbance stages include the seismic activity stage, the foundation settlement stage, and the fault sliding stage. The disturbances are then sorted from high to low intensity to generate a geological disturbance sequence. Based on the climate environment information, the climate erosion stages are divided, including the high-temperature weathering stage, the freeze-thaw cycle stage, and the dry-wet alternation stage, and are sorted from high to low according to the erosion intensity to generate a climate erosion sequence. The hydrological load sequence, geological disturbance sequence, and climate erosion sequence are arranged and combined in three dimensions. The results of the arrangement and combination are weighted and sorted based on the intensity of load, disturbance, and erosion to generate a comprehensive driving load sequence. Based on the dam monitoring database, a comprehensive sensitivity threshold is set, and the comprehensive driving load sequence is evaluated for aging driving load using the comprehensive sensitivity threshold.

3. The dam aging assessment method based on GNSS data time-series characteristic analysis as described in claim 2, characterized in that, The step of evaluating the aging-driven load of the comprehensive driving load sequence using the comprehensive sensitivity threshold includes: Traverse the comprehensive driving load sequence to determine whether there is a load item that matches the comprehensive sensitivity threshold, wherein the load item is a combination of single environmental effects in a three-dimensional permutation and combination; If it exists, the comprehensive sensitivity threshold is marked on the corresponding load item, and the proportion of load items higher than the comprehensive sensitivity threshold and the proportion of load items lower than the comprehensive sensitivity threshold are determined. The aging-driven load is evaluated based on the proportion. If none exists, then select any of the load items and determine the relationship between the combined effect intensity of the load item and the combined sensitivity threshold. If the overall sensitivity threshold is higher, the overall driving load sequence is marked as a high-risk load sequence; if the overall sensitivity threshold is lower, the overall driving load sequence is marked as a low-risk load sequence.

4. The dam aging assessment method based on GNSS data time series characteristic analysis as described in claim 3, characterized in that, The step of evaluating the aging-driven load based on the stated proportion includes: The design reference period of the dam is obtained, and the aging rate of the dam under each load item is calculated based on the comprehensive sensitivity threshold and proportion. The durability performance degradation trend of the dam structure is evaluated in conjunction with the design reference period. A short-term impact recovery threshold is set, which is the critical value of the deformation recovery ability of the dam material under short-term high load. The dam is subjected to impact simulation using a load term higher than the comprehensive sensitivity threshold within a unit time to obtain the dam's resistance to sudden damage. The aging-driven load is evaluated based on the durability degradation trend and resistance to sudden damage performance.

5. The dam aging assessment method based on GNSS data time-series characteristic analysis as described in claim 1, characterized in that, The establishment of the dam monitoring database includes: Collect historical maintenance and monitoring data of the dam, including structural design schemes of dams of different types and construction years and corresponding GNSS monitoring data. The time series characteristic parameters of the GNSS data include cumulative displacement, displacement rate, periodic fluctuation amplitude and abrupt displacement. The historical maintenance and monitoring data information is classified according to the basic structural parameters, geographical environment information, dam structural design information, and GNSS data time series characteristic parameters. The classification includes classification by dam type, classification by construction year, and classification by geographical region. A multi-dimensional data index entry is established based on the classification results. Structural design subsets are constructed according to each dam's structural design scheme, and a GNSS time-series feature tracking chain is established for each structural design scheme, so that the aging simulation space can call historical time-series data.

6. The dam aging assessment method based on GNSS data time series characteristic analysis as described in claim 5, characterized in that, The historical time-series data available for the aging simulation space to access includes: The hydrological, geological, and climatic environmental information of the dam is determined. Historical environmental parameters from the same period are matched by indexing the dam monitoring database, and environmental simulation factors are constructed. These environmental simulation factors are digital quantitative representations of the geographic environmental information in the aging simulation space. The basic structural parameters and structural design information of the dam are determined. The structural parameters and design parameters of similar dam types are matched by indexing the dam monitoring database. The accelerated aging time is divided into stages. Each stage is set with a different time acceleration factor according to the aging rate characteristics. The time acceleration factor is positively correlated with the rate of change of GNSS time series characteristic parameters. Based on the determined current geographical environment information and current dam structure information, the dam monitoring database is matched and the evolution law of the corresponding GNSS data time series characteristics is tracked to provide time series characteristic evolution model support for the aging simulation results.

7. The dam aging assessment method based on GNSS data time-series characteristic analysis as described in claim 1, characterized in that, The dynamic simulation of several aging stages of the dam based on the accelerated aging time includes: Set a time acceleration factor to define the conversion ratio between the length of time elapsed during the simulated aging process and the actual running time, thereby accelerating the simulation process to quickly reproduce the long-term aging effect. Based on the actual geographical environment information, basic structural parameters and GNSS data time series characteristic parameters of the dam, several specific aging stages are defined. The aging stages include the initial micro-deformation stage, the intermediate crack development stage and the late structural damage stage, and each aging stage corresponds to a time acceleration factor. In each defined aging stage, corresponding hydrological loads, geological disturbances, and climate erosion parameters are loaded, and the aging process is dynamically simulated by combining the dynamic changes of GNSS data time series characteristic parameters. By combining the simulation results of all the aging stages, the evolution trend of GNSS time series characteristics of the dam in the assumed whole life cycle is evaluated, and the aging simulation results are obtained.

8. The dam aging assessment method based on GNSS data time-series characteristic analysis as described in claim 1 or 7, characterized in that, The aging assessment results of the dam obtained through aging simulation include: Deep learning training is performed on the GNSS data time series feature parameters in the dam monitoring database to obtain the mapping relationship between the basic structure parameter information, geographical environment information, dam structure design information and the GNSS data time series feature parameters, as well as the nonlinear mapping relationship between the basic structure parameter information, geographical environment information, dam structure design information and the GNSS data time series feature parameters after coupling. Set a real-time monitoring cycle, collect GNSS monitoring data of the dam in operation in real time, extract real-time time series characteristic parameters and enter them into the dam monitoring database, and dynamically update the dam monitoring database according to the real-time monitoring cycle, and adjust the mapping relationship synchronously. Based on the mapping relationship, the GNSS time-series characteristic parameters in the aging simulation results are trend-predicted, and the prediction results are evaluated in conjunction with the dam aging level classification standard to obtain the dam aging assessment results, which include aging level, key weak areas and remaining safe operating years.

9. A dam aging assessment system based on GNSS data time-series characteristic analysis, characterized in that, The device includes: a memory, a processor, and a dam aging assessment program based on GNSS data time-series feature analysis stored in the memory and executable on the processor, the dam aging assessment program based on GNSS data time-series feature analysis being configured to implement the steps of the dam aging assessment method based on GNSS data time-series feature analysis as described in any one of claims 1 to 8.

10. A medium, characterized in that, The medium stores a dam aging assessment program based on GNSS data time-series feature analysis. When the dam aging assessment program based on GNSS data time-series feature analysis is executed by the processor, it implements the steps of the dam aging assessment method based on GNSS data time-series feature analysis as described in any one of claims 1 to 8.

Images