Bridge slope geological disaster early warning system and method based on multi-source data fusion

Abstract

The invention discloses a bridge slope geological disaster early warning system and method based on multi-source data fusion, and belongs to the technical field of bridge slope geological disaster early warning, and the system comprises an obtaining module which is used for obtaining the monitoring data of a to-be-detected bridge slope; the processing module is used for taking abnormal data existing in the monitoring data as early abnormal data; the prediction module is used for predicting a first probability of landslide of the to-be-detected bridge slope based on the early abnormal data, predicting a second probability of landslide based on the monitoring data, and determining a landslide risk level of landslide based on the first probability and the second probability; and the early warning module is used for carrying out early warning analysis on a landslide-debris flow-bridge damage disaster chain and carrying out early warning by utilizing an early warning model if the landslide risk grade is greater than or equal to the first risk grade. The system can accurately predict whether a landslide disaster occurs on a bridge slope or not so as to timely and accurately carry out early warning prompt of a landslide-debris flow-bridge damage disaster chain.

Description

Technical Field

[0001] The present invention belongs to the technical field of geological disaster warning for bridge slopes, and particularly relates to a geological disaster warning system and method for bridge slopes based on multi-source data fusion. Background Art

[0002] With the continuous advancement of infrastructure construction, highway bridge construction has extended from the original plain and gentle hilly areas to the current hilly and mountainous areas. The terrain in hilly areas is steep and slopes are widely distributed. Topography and geological conditions are the main influencing factors in highway design in hilly mountainous areas. Considering drainage conditions and highway grade factors comprehensively, bridge structures are often selected to cross complex terrains located in the sliding zone. Therefore, large-scale excavated slopes will appear on both sides of the bridge. At this time, the stability of the bridge slope is closely related to the safety of the bridge. When a landslide occurs on the bridge slope, it is extremely easy to trigger debris flow, which will cause serious damage to the bridge, that is, a landslide-debris flow-bridge damage disaster chain will occur, resulting in relatively large disaster losses, such as bridge property and safety losses. Therefore, predicting in advance whether a landslide disaster will occur on the bridge slope is beneficial to timely and accurately give early warnings of the landslide-debris flow-bridge damage disaster chain. However, the existing technology usually draws the historical landslide curves corresponding to each control factor in historical cases of existing landslides, then draws the current landslide curve corresponding to the control factor of the slope to be measured, and compares it with the historical landslide curve to predict the probability of a landslide occurring on the slope to be measured. However, the accuracy of predicting whether a landslide disaster will occur on the bridge slope in this way is relatively low, resulting in difficulty in timely and accurately giving early warnings of the landslide-debris flow-bridge damage disaster chain.

[0003] It should be noted that the information disclosed in the above background art section is only used to enhance the understanding of the background of the present application, and thus may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0004] To have a basic understanding of some aspects of the disclosed embodiments, a simple summary is given below. The summary is not a general review, nor is it intended to identify key / important constituent elements or delineate the protection scope of these embodiments, but rather serves as a preamble to the subsequent detailed description.

[0005] The embodiments of the present disclosure provide a geological disaster warning system and method for bridge slopes based on multi-source data fusion, so as to be able to accurately predict whether a landslide disaster will occur on the bridge slope, thereby being able to timely and accurately give early warnings of the landslide-debris flow-bridge damage disaster chain.

[0006] In some embodiments, the bridge slope geological disaster early warning system based on multi-source data fusion includes: an acquisition module for acquiring monitoring data of a to-be-detected bridge slope; wherein the monitoring data includes geological terrain data, meteorological and hydrological data, engineering activity data, and displacement and deformation data; a processing module for determining whether there is abnormal data based on the monitoring data, and if so, taking the abnormal data as preliminary abnormal data; a prediction module for predicting a first probability of landslide occurrence of the to-be-detected bridge slope based on the preliminary abnormal data, and predicting a second probability of landslide occurrence of the to-be-detected bridge slope based on the monitoring data, and determining a landslide risk level of the to-be-detected bridge slope based on the first probability and the second probability; and an early warning module for, if the landslide risk level is greater than or equal to a first risk level, performing a disaster chain early warning analysis of landslide-debris flow-bridge damage using a preset early warning model, obtaining an early warning analysis result, and performing an early warning based on the early warning analysis result.

[0007] In some embodiments, the bridge slope geological disaster early warning method based on multi-source data fusion includes: acquiring monitoring data of a to-be-detected bridge slope; wherein the monitoring data includes geological terrain data, meteorological and hydrological data, engineering activity data, and displacement and deformation data; determining whether there is abnormal data based on the monitoring data, and if so, taking the abnormal data as preliminary abnormal data; predicting a first probability of landslide occurrence of the to-be-detected bridge slope based on the preliminary abnormal data, and predicting a second probability of landslide occurrence of the to-be-detected bridge slope based on the monitoring data, and determining a landslide risk level of the to-be-detected bridge slope based on the first probability and the second probability; and if the landslide risk level is greater than or equal to a first risk level, performing a disaster chain early warning analysis of landslide-debris flow-bridge damage using a preset early warning model, obtaining an early warning analysis result, and performing an early warning based on the early warning analysis result.

[0008] The beneficial effects of the present invention are as follows:

[0009] The monitoring data of the bridge slope to be measured is obtained through an acquisition module. The abnormal data in the monitoring data is used as preliminary abnormal data by a processing module. The preliminary abnormal data is used to reflect the local abnormality of the bridge slope to be measured, while the monitoring data reflects the overall state of the bridge slope. In this way, a prediction module uses the preliminary data to predict the first probability of landslide occurrence on the bridge slope to be measured, uses the monitoring data to predict the second probability of landslide occurrence on the bridge slope to be measured, and then determines the landslide risk level of the bridge slope to be measured based on the first probability and the second probability, so as to realize the prediction of whether the bridge slope to be measured will have a landslide based on the local abnormality and the overall state of the bridge slope to be measured, so as to be able to more accurately predict whether the bridge slope will have a landslide disaster. When the landslide risk level is greater than or equal to the first risk level, it indicates that the probability of landslide occurrence on the bridge slope to be measured is relatively large at this time. At this time, an early warning model is used to conduct early warning analysis of the landslide-debris flow-bridge damage disaster chain, so as to be able to give timely and accurate early warning prompts for the landslide-debris flow-bridge damage disaster chain.

[0010] The above general description and the description in the following text are only exemplary and explanatory, and are not used to limit the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] One or more embodiments are exemplarily illustrated by corresponding drawings. These exemplary illustrations and the drawings do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are shown as similar elements. The drawings do not constitute a scale limitation, and in which:

[0012] Figure 1 is a schematic structural diagram of a bridge slope geological disaster early warning system based on multi-source data fusion provided by the present invention;

[0013] Figure 2 is a flowchart of a method for early warning of bridge slope geological disasters based on multi-source data fusion provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0014] In order to be able to understand the features and technical content of the embodiments of the present disclosure in more detail, the implementation of the embodiments of the present disclosure will be described in detail below with reference to the drawings. The attached drawings are only for reference and explanation, and are not used to limit the embodiments of the present disclosure. In the following technical description, for the sake of explanation, numerous details are provided to give a full understanding of the disclosed embodiments. However, one or more embodiments can still be implemented without these details. In other cases, well-known structures and devices may be shown in a simplified manner to simplify the drawings.

[0015] In the description, claims, and above-mentioned accompanying drawings of the embodiments of the present disclosure, terms such as "first" and "second" are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so as to implement the embodiments of the present disclosure described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion.

[0016] Unless otherwise specified, the term "plurality" means two or more.

[0017] In the embodiments of the present disclosure, the character " / " indicates that the objects before and after are in an "or" relationship. For example, A / B means: A or B.

[0018] The term "and / or" is a description of the association relationship of an object, indicating that three relationships can exist. For example, A and / or B means: A or B, or, A and B these three relationships.

[0019] The term "corresponding" may refer to an association relationship or a binding relationship. A corresponding to B means that there is an association relationship or a binding relationship between A and B.

[0020] Combined Figure 1 As shown, the embodiments of the present disclosure provide a bridge slope geological disaster early warning system based on multi-source data fusion, including an acquisition module, a processing module, a prediction module, and an early warning module. Among them, the acquisition module is used to acquire the monitoring data of the bridge slope to be measured; among them, the acquisition module is used to acquire the monitoring data of the bridge slope to be measured; among them, the monitoring data includes geological terrain data, meteorological and hydrological data, engineering activity data, and displacement and deformation data. The processing module is used to determine whether there is abnormal data based on the monitoring data, and if so, use the abnormal data as the preliminary abnormal data. The prediction module is used to predict the first probability of a landslide occurring on the bridge slope to be measured based on the preliminary abnormal data, and predict the second probability of a landslide occurring on the bridge slope to be measured based on the monitoring data, and determine the landslide risk level of the bridge slope to be measured based on the first probability and the second probability. The early warning module is used to, if the landslide risk level is greater than or equal to the first risk level, perform disaster chain early warning analysis of landslide-debris flow-bridge damage using a preset early warning model, obtain the early warning analysis result, and perform early warning based on the early warning analysis result.

[0021] Adopt the bridge slope geological disaster early warning system based on multi-source data fusion provided by the embodiments of the present disclosure. The monitoring data of the bridge slope to be measured is obtained through the acquisition module, and the abnormal data in the monitoring data is used as the preliminary abnormal data through the processing module. The preliminary abnormal data is used to reflect the local abnormality of the bridge slope to be measured, while the monitoring data reflects the overall state of the bridge slope. In this way, through the prediction module, the first probability of landslide occurring on the bridge slope to be measured is predicted using the preliminary data, the second probability of landslide occurring on the bridge slope to be measured is predicted using the monitoring data, and then the landslide risk level of the bridge slope to be measured is determined based on the first probability and the second probability, so as to realize the prediction of whether the bridge slope to be measured will have a landslide based on the local abnormality and the overall state of the bridge slope to be measured, so as to be able to accurately predict whether the bridge slope will have a landslide disaster. When the landslide risk level is greater than or equal to the first risk level, it indicates that the probability of landslide occurring on the bridge slope to be measured is relatively large at this time. At this time, the early warning module is used to conduct early warning analysis of the landslide-debris flow-bridge damage disaster chain, so as to be able to timely and accurately give early warning prompts for the landslide-debris flow-bridge damage disaster chain.

[0022] In some embodiments, the geological and topographical data includes rock and soil properties, geological structures, slope gradient, slope length, vegetation cover, and soil moisture saturation; for example, the rock and soil properties and geological structures of the slope of the bridge to be measured can be obtained through geological exploration, the slope gradient and slope length of the slope of the bridge to be measured can be obtained by using topographic maps or digital elevation models (DEMs) or lidar, etc., and the vegetation cover of the slope of the bridge to be measured can be monitored through remote sensing or on-site surveys, etc. The meteorological and hydrological data includes rainfall information (such as rainfall amount, rainfall intensity), groundwater information, and freeze-thaw cycle information. The engineering activities include engineering construction activities within a first preset range centered on the slope of the bridge, and this first preset range represents the range where the engineering construction activities have an obvious impact on the stability of the bridge slope. It can be understood that the intensity of the engineering construction activities has different degrees of influence on the stability of the bridge slope. Therefore, two sub-regions can be set, such as a first sub-region centered on the slope of the bridge and a second sub-region centered on the slope of the bridge (the second sub-region includes and is larger than the first sub-region); then determine the intensity of the engineering construction activities (such as two intensities: high and low), and regard all the engineering construction activities within the first sub-region as the engineering activities that have an impact on the stability of the bridge slope, and regard the engineering construction activities with high intensity within the second sub-region as the engineering activities that have an impact on the stability of the bridge slope. The displacement and deformation data includes surface displacement (such as horizontal and vertical displacements), deep displacement (such as slip surface displacement and layered settlement), crack conditions (width, length, and trend changes of cracks on the slope surface), and overall deformation (three-dimensional deformation field, such as composite deformations like inclination and torsion); the surface displacement can be obtained by combining GNSS and total station, the deep displacement can be obtained by combining inclinometers and settlement gauges, the crack conditions can be obtained by combining electronic crack meters and photogrammetry, and the overall deformation can be obtained by combining three-dimensional laser scanning and InSAR (satellite remote sensing). That is, in the embodiments of the present disclosure, the monitoring data is multi-source data, including geological and topographical data, meteorological and hydrological data, engineering activity data, and displacement and deformation data. For example, multi-source data such as geological exploration data, topographic map data, remote sensing data, meteorological and hydrological monitoring data, construction project information research data, GNSS data, total station monitoring data, inclinometer monitoring data, settlement gauge monitoring data, three-dimensional laser scanning data, and InSAR (satellite remote sensing) data.

[0023] In some embodiments, the early abnormal data represents the precursor monitoring data before a landslide occurs on the slope of the bridge to be measured. For the geotechnical properties and geological structures in the geological and topographical data, they are relatively stable and usually do not change. Engineering activities are dominated by human activities, and it is difficult to quantify whether they are abnormal. Therefore, when determining whether there is abnormal data in the monitoring data, the geotechnical properties and geological structures in the geological and topographical data, as well as the engineering activity data, can be removed. For the remaining monitoring data, its changes should have its own characteristics. When there are changes different from its own characteristics, it indicates that it may be abnormal. For example, the processing module is specifically used to: use the data in the monitoring data except for the geotechnical properties, geological structures, and engineering activity data as the target monitoring data. Divide the target monitoring data into multiple consecutive time windows, and calculate the dynamic change data of the slope of the bridge to be measured in each time window respectively. The dynamic change data includes: slope change rate, slope length change rate, vegetation coverage rate, soil moisture saturation, rainfall change rate, groundwater level change range, freezing depth and moisture content of freeze-thaw cycles, surface displacement change rate, deep displacement change rate, crack condition change rate, overall deformation change rate, and set the threshold conditions corresponding to each kind of dynamic change data. For each kind of dynamic change data, when it meets the corresponding threshold condition in any one time window, it is considered that the target monitoring data corresponding to this kind of dynamic change data is the early abnormal data, and the time window that meets the threshold condition is determined as the duration of the abnormality of this early abnormal data. The total duration corresponding to the monitoring data is used as the total monitoring duration, and the sum of the durations of the abnormalities of the early abnormal data is used as the cumulative abnormal duration.

[0024] Preferably, the prediction module includes: a first sub-module and a second sub-module. Among them, the first sub-module is used to calculate the sub-probability of each kind of early abnormal data causing a landslide on the slope of the bridge to be measured respectively. The second sub-module is used to comprehensively determine the first probability based on each sub-probability.

[0025] In this way, by calculating the sub-probability corresponding to each kind of early abnormal data respectively and then comprehensively determining the first probability of causing a landslide on the slope of the bridge to be measured, the influence of various early abnormal data on the slope of the bridge to be measured can be determined more accurately.

[0026] Preferably, the first sub-module includes: a first calculation unit and a second calculation unit. Among them, the first calculation unit is used to calculate the instability probability of the natural instability of the slope of the bridge to be measured, calculate the correlation degree between each kind of early abnormal data and the slope of the bridge to be measured, and the probability of the occurrence of each kind of early abnormal data. The second calculation unit is used to, for each kind of early abnormal data, calculate the sub-probability of each kind of early abnormal data causing a landslide on the slope of the bridge to be measured based on the instability probability, the correlation degree, and the probability of the occurrence of the early abnormal data by using the Bayesian formula, as follows:

[0027]

[0028] wherein, P(A / B i ) is the sub - probability that the landslide of the bridge slope to be measured is caused by the i - th type of prior abnormal data, P(B i / A) is the correlation degree between the i - th type of prior abnormal data and the bridge slope to be measured, P(A) is the instability probability, and P(B i ) is the probability of the occurrence of the i - th type of prior abnormal data. Among them, natural instability represents the occurrence of a landslide under natural conditions.

[0029] In some embodiments, the natural conditions can be regarded as the situation where there is no obvious inducement for the landslide of the slope. The obvious inducements include extreme rainfall events and / or earthquake events. That is, in the embodiments of the present disclosure, natural instability represents the occurrence of a landslide without extreme rainfall events and without earthquake events. An extreme rainfall event represents that the rainfall amount ≥ extreme rainfall threshold, and / or the rainfall intensity ≥ extreme intensity threshold. For example, if slope A has a landslide under the influence of an extreme rainfall event or an earthquake event, then slope A is non - naturally unstable. If slope B has a landslide without the influence of extreme rainfall and earthquake events, then slope B is naturally unstable. Since the influence of micro - seismic (such as a magnitude - 1 earthquake) on the bridge slope is small and it is difficult to directly cause the bridge slope to have a landslide, the earthquake magnitude condition can also be set as an obvious inducement according to the actual situation. For example, an earthquake event with a magnitude higher than the preset earthquake magnitude is used as an obvious inducement.

[0030] In this way, by calculating the instability probability, the correlation degree corresponding to each type of prior abnormal data, and the probability of the occurrence of each type of prior abnormal data, and then using the Bayes formula for calculation, it is possible to more accurately obtain the sub - probability that each type of prior abnormal data causes the bridge slope to have a landslide.

[0031] Preferably, the first calculation unit is specifically configured to: obtain the total number of slopes in the preset monitoring area and the number of slopes that have natural instability; based on the total number of slopes in the preset monitoring area and the number of slopes that have natural instability in the preset monitoring area, use the classical probability formula to calculate the probability that the slopes in the preset monitoring area have natural instability, as follows: wherein, P(C) is the probability that the slopes in the preset monitoring area have natural instability, m is the total number of slopes in the preset monitoring area, and n is the number of slopes that have natural instability in the preset monitoring area; the probability that the slopes in the preset monitoring area have natural instability is used as the instability probability that the bridge slope to be measured has natural instability.

[0032] In this way, by setting the monitoring area, obtaining the total number of slopes within the monitoring area and the number of slopes that have undergone natural instability, and using the classical probability formula to calculate the probability of natural instability of the slopes within the preset monitoring area, and taking it as the instability probability of the slopes of the bridge to be measured undergoing natural instability, the instability probability of the slopes of the bridge to be measured undergoing natural instability can be obtained more accurately.

[0033] It can be understood that the monitoring area is an area with a relatively large number of slopes, so as to obtain sufficient sample data (i.e., the total number of slopes, the number of slopes that have undergone natural instability), thereby improving the calculation of the probability of natural instability of the slopes within this area.

[0034] In this embodiment, due to the different climatic conditions in each region, the extreme rainfall threshold and extreme intensity threshold will also vary. For the preset monitoring area, multiple rainfall events occurring within this area (the rainfall events include rainfall amount and rainfall intensity) can be randomly selected, divided according to seasons or rainfall pattern characteristics, to obtain different categories (such as seasons: spring, summer, autumn, and winter; rainfall pattern characteristics: multi-rainfall periods, few-rainfall periods), and then the extreme rainfall threshold and extreme intensity threshold corresponding to each category are determined respectively. Among them, the extreme rainfall threshold and extreme intensity threshold corresponding to each category can be obtained respectively through the following methods:

[0035] For the extreme rainfall threshold of each category, multiple rainfall events are selected, and the variance of the overall rainfall amount is determined using the T-distribution: Among them, is the mean value of the selected rainfall amounts, X i is the i-th rainfall amount, n is the number of selected rainfall events, S is the sample standard deviation, t is the variance of the overall rainfall amount, and μ is the mean value of the overall rainfall amount. Then, a confidence interval is constructed using the confidence level and the variance of the overall rainfall amount, and the upper limit of this confidence interval is taken as the initial rainfall threshold: Among them, α is the significance level, then (1 - α) is the preset confidence level, represents the t-value corresponding to the right-tail area in the t-distribution with n - 1 degrees of freedom corresponding to, represents the confidence width, and the confidence width describes the distribution interval of the acceptance region, is the lower limit of the confidence interval, is the upper limit of the confidence interval; then, the product of the initial rainfall threshold and a preset parameter is taken as the extreme rainfall threshold, and this first preset parameter is greater than 1 (the extreme rainfall threshold should be significantly greater than the initial rainfall threshold), and its specific value can be determined according to the actual situation or expert experience. Similarly, the extreme intensity threshold is determined, and the details will not be elaborated here.

[0036] In some embodiments, when calculating the probability of natural instability of slopes within a preset monitoring area, first, the slopes within the monitoring area that have experienced landslides are regarded as target slopes. It is determined whether there are extreme rainfall events or earthquake events within a preset historical duration (such as the previous 10 days) before the landslide occurred on the target slope. If there are, then the target slope is regarded as a slope with non-natural instability; if not, then the target slope is regarded as a slope with natural instability. For example, for each target slope, rainfall events and earthquake events within the preset historical duration before the landslide occurred are obtained. Based on the rainfall event, the category it belongs to (such as spring) is determined, and then based on the extreme rainfall threshold and extreme intensity threshold of this category, it is determined whether the rainfall event is an extreme rainfall event. If it belongs to an extreme rainfall event, then the slope triggered by this extreme rainfall event to have a landslide is regarded as a slope with non-natural instability, and the difference between the number of landslides that occurred within this monitoring area and the number of slopes with non-natural instability is regarded as the number of natural instabilities (that is, the number of slopes with natural instability). Similarly, based on the earthquake events within the preset historical duration before the landslide occurred, it is determined whether the target slope is a slope with non-natural instability or a slope with natural instability. A slope with non-natural instability represents a slope triggered by extreme rainfall events and / or earthquake events to have a landslide. A slope with natural instability represents a slope that has a landslide in the absence of extreme rainfall events and earthquake events. For example, Area A is the monitoring area. There are a total of 100 slopes within Area A, and 8 slopes have experienced landslides. Among them, 2 slopes are slopes with non-natural instability, then the number of slopes with natural instability is 6. The probability of natural instability of the slopes in the monitoring area of Area A = (8 - 2) / 100.

[0037] Preferably, the preset monitoring area is determined based on the characteristics of the bridge to be measured, and the characteristics include at least one of administrative characteristics, scope characteristics, mountain characteristics, and geotechnical properties. In some embodiments, the preset monitoring area is determined based on the administrative characteristics of the bridge to be measured: the area pointed to by the superior governing unit to which the slope of the bridge to be measured belongs is used as the preset monitoring area. For example, if the slope of the bridge to be measured is located in Town A, then the preset area can be the area pointed to by the superior governing unit to which Town A belongs, that is, Province A, City A, County A, etc. In some other embodiments, the preset monitoring area is determined based on the scope characteristics of the bridge to be measured: the area within a second preset range centered on the slope of the bridge to be measured is used as the preset monitoring area. For example, the second preset range is within a few hundred kilometers. In some other embodiments, the preset monitoring area is determined based on the mountain characteristics of the bridge to be measured: the mountain to which the slope of the measured bridge belongs is used as the preset monitoring area. For example, if the slope of the bridge to be measured is located at a certain position on Mountain A, then Mountain A is used as the preset monitoring area. In some other embodiments, the preset monitoring area is determined based on the geotechnical characteristics of the bridge to be measured: the area with the same or similar geotechnical properties as the slope of the bridge to be measured is used as the preset monitoring area. Geotechnical properties include rock strength, soil properties, and geotechnical structure.

[0038] Preferably, the correlation degree between the early abnormal data and the bridge slope to be measured characterizes the influence degree of the early abnormal data on the landslide of the bridge slope; that is, the greater the correlation degree, the easier it is for the early abnormal data to cause the landslide of the bridge slope. In some embodiments, the influence degree of various early abnormal data on the landslide of the bridge slope can be determined by the expert experience method, and this influence degree is used as the corresponding correlation degree.

[0039] Preferably, the first calculation unit is configured to calculate the probability of occurrence of each type of early abnormal data in the following manner: respectively obtain the total monitoring duration of each type of early abnormal data and its cumulative abnormal duration when an abnormality occurs; for each type of early abnormal data, based on the total monitoring duration and the cumulative abnormal duration, calculate the probability of occurrence of the early abnormal data through the following formula: where P(B i ) is the probability of occurrence of the i-th type of early abnormal data, t i is the cumulative abnormal duration when the i-th type of early abnormal data shows an abnormality, and T i is the total monitoring duration of the i-th type of early abnormal data.

[0040] In this way, for each type of early abnormal data, based on calculating the proportion of the cumulative abnormal duration in its total monitoring duration, the probability of its occurrence can be determined more reasonably and accurately.

[0041] Preferably, the second sub-module is specifically configured to comprehensively determine the first probability based on each sub-probability in the following manner: P z = 1 - (1 - P1)·(1 - P2)·(1 - P3)...·(1 - P i ). Where P z is the first probability, and P i is the sub-probability that the i-th type of early abnormal data causes the landslide of the bridge slope to be measured.

[0042] In some embodiments, the prediction module may also use a combination of a knowledge graph and a graph convolutional neural network to predict a second probability of landslide occurrence on the slope of the bridge to be measured based on the monitoring data. For example, according to the node information of the slope of the bridge to be measured, a knowledge graph ontology model is constructed, and then a landslide disaster knowledge graph is constructed based on multiple historical landslide events and the ontology model. The historical landslide events include the corresponding monitoring data, and the knowledge graph includes geological and topographical nodes, meteorological and hydrological nodes, engineering activity nodes, and displacement and deformation nodes. Then, a graph convolutional neural network is constructed. For each historical landslide event, the feature vector corresponding to the monitoring data is input into the graph convolutional neural network for training to obtain a corresponding landslide prediction model. Then, based on the feature vector corresponding to the monitoring data of the slope of the bridge to be measured, the second prediction probability can be obtained by using the landslide prediction model. The feature vector can be the static information of the monitoring data (such as rainfall) or the dynamic change information (such as the displacement change rate). In this embodiment, if the monitoring duration corresponding to the monitoring data is relatively long, the latest monitoring data in the monitoring data can be selected as the input data of the landslide prediction model. The latest time can be the time data of the latest 1 day in the monitoring data.

[0043] In other embodiments, the prediction module may also use a linear regression network to predict a second probability of landslide occurrence on the slope of the bridge to be measured based on the monitoring data. For example, a prediction function is constructed and trained based on multiple historical landslide events to obtain a target prediction function: F(x) = α0 + α1x1 + α2x2 + α3x3 +... + α n x n , where x = {x1, x2, x3,... x n} is the feature vector corresponding to the monitoring data, α0 is the intercept term, and α1, α2, α3,... α n are the regression coefficients, indicating the influence degree of each feature vector on the prediction result. The prediction function obtains the second probability of landslide occurrence on the slope of the bridge to be measured by summing the products of each feature vector and the corresponding regression coefficient and then adding the intercept term. The function training process is a relatively common technology and will not be elaborated here.

[0044] Preferably, the prediction module includes a third sub-module; the third sub-module is specifically configured to: respectively determine a first level corresponding to the first probability and a second level corresponding to the second probability; if either the first level or the second level is greater than or equal to the second risk level, determine the highest level among the first level and the second level as the landslide risk level of the slope of the bridge to be measured; if both the first level and the second level are less than the second risk level, determine the landslide risk level of the slope of the bridge to be measured based on the first level, the first preset weight corresponding to the first level, the second level, and the second preset weight corresponding to the second level; where the second risk level is greater than the first risk level.

[0045] In this way, by mapping the first probability and the second probability to the corresponding levels, if the first level is greater than or equal to the second risk level, it indicates that based on the prior abnormal data, the slope of the bridge to be measured is likely to have a landslide. If the second level is greater than or equal to the second risk level, it indicates that based on the monitoring data, the slope of the bridge to be measured is likely to have a landslide. At this time, the highest level among the first level and the second level is determined as the landslide risk level of the slope of the bridge to be measured, so as to be able to predict more accurately whether the slope of the bridge to be measured will have a landslide. If both the first level and the second level are less than the second risk level, it means that it cannot be directly determined whether the slope of the bridge to be measured will have a landslide only based on the prior abnormal data or the slope of the bridge to be measured. At this time, the landslide risk level of the slope of the bridge to be measured is determined by combining the first preset weight, the second preset weight, the first level and the second level, so as to be able to determine more accurately whether the slope of the bridge to be measured will have a landslide.

[0046] Preferably, the third sub-module is specifically configured to determine the first level corresponding to the first probability and the second level corresponding to the second probability respectively in the following manner: perform a look-up operation in the preset data table using the first probability and the second probability respectively, and find the level corresponding to the first probability as the first level, and find the level corresponding to the second probability as the second level. Among them, the corresponding relationship between the first probability, the second probability and the level is stored in the preset data table.

[0047] Preferably, the third sub-module is specifically configured to, if both the first level and the second level are less than the second risk level, determine the landslide risk level of the slope of the bridge to be measured based on the first level, the first preset weight corresponding to the first level, the second level and the second preset weight corresponding to the second level: perform a look-up operation based on the first level and the second level respectively, and find the mapped value corresponding to the first level as the first target value, and find the mapped value corresponding to the second level as the second target value. Calculate the landslide risk value through the following formula, landslide risk value = first target value * first preset weight + second target value * second preset weight. If the landslide risk value is greater than or equal to the preset risk value, the landslide risk level of the slope of the bridge to be measured is determined as the first risk level.

[0048] In some embodiments, the first preset weight is 0.6 and the second preset weight is 0.4. Table 1 is an example table of the preset data table. The first risk level is level three, the second risk level is level four, and the preset risk value is 2a. Assuming P1 is 28% and P2 is 20%, then the first level is level four and the first level is level three, so the landslide risk level is level four. Assuming P1 is 15% and P2 is 20%, then the first level is level two, the corresponding first target value is 2a, the second level is level three, and the corresponding second target value is 3a. Then the landslide risk value is 2.4a (2a * 0.6 + 3a * 0.4). Since this landslide risk value is greater than the preset risk value, the landslide risk level of the bridge slope to be measured is determined as the first risk level, that is, level three.

[0049] Table 1

[0050] <![CDATA[First probability P1]]> <![CDATA[First probability P2]]> Level Mapping value <![CDATA[P1 < 15%]]> <![CDATA[P2 < 15%]]> First level a <![CDATA[15% ≤ P1 < 20%]]> <![CDATA[15% ≤ P2 < 20%]]> Second level 2a <![CDATA[20% ≤ P1 < 25%]]> <![CDATA[20% ≤ P2 < 25%]]> Third level 3a <![CDATA[25% ≤ P1]]> <![CDATA[25% ≤ P2]]> Fourth level None

[0051] Preferably, a geological disaster early warning system for bridge slopes based on multi-source data fusion further includes a model construction module; the model construction module includes: a node sub-module, a determination sub-module, a case sub-module, a calculation sub-module, and a model sub-module. Among them, the node sub-module is used to obtain nodes related to the "landslide-debris flow-bridge damage" disaster chain; among them, the nodes include three types of nodes: disaster-causing factors, disaster-bearing environments, and disaster events. The disaster-causing factor nodes include landslide body characteristics and sliding paths. The disaster-bearing environment nodes include rainfall intensity, longitudinal gradient of the gully, debris flow fluid characteristics, relative position of the bridge and the gully, and bridge structure. The disaster event nodes include debris flow grade and bridge damage grade. The determination sub-module is used to determine the influence relationship between each node based on the disaster-causing mechanism and chain transfer law of the "landslide-debris flow-bridge damage" disaster chain, and determine the directed edge information between each node based on the influence relationship. The case sub-module is used to obtain corresponding multiple historical cases based on the "landslide-debris flow-bridge damage" disaster chain; the historical cases include node variables corresponding to each node. The calculation sub-module is used to calculate the conditional probability between each node by using the node variables of each node in multiple historical cases. The model sub-module is used to construct a Bayesian network model based on each node, the directed edge information between each node, and the conditional probability between each node, and determine the constructed Bayesian network model as the early warning model. Among them, the node variable represents the numerical information corresponding to the node in the historical case.

[0052] In this way, by obtaining the nodes related to the "landslide-debris flow-bridge damage" disaster chain, determining the influence relationship between each node, and then calculating the conditional probability between each node by using the node variables of each node in each historical case, it is convenient to construct a Bayesian network model corresponding to the "landslide-debris flow-bridge damage" disaster chain, so as to be able to accurately early warn of the disasters caused by bridge slopes.

[0053] In some embodiments, the landslide body characteristics characterize the landslide intensity, including the landslide volume, the type of rock and soil (such as cohesive soil, sandy soil, gravel and bedrock blocks), and the slope. The sliding paths include valley-constrained type (the landslide body moves along the existing valley or river channel, restricted by the terrain on both sides, making the path relatively concentrated), fan-shaped diffusion type (the landslide body spreads in a fan shape in the flat area after rushing out of the valley mouth, with a wide path), straight collapse type (the landslide body slides straight down along a single steep slope without obvious valley guidance), and jumping collision type (the landslide body bounces and collides on the steep cliff or stepped terrain, with a discontinuous path); the debris flow fluid characteristics characterize the debris flow intensity, including the capacity, the flow velocity, and the solid particle composition.

[0054] Preferably, the determining sub-module is specifically configured to: if the occurrence state of one node may affect the occurrence state of another node, then for these two nodes, the node that affects the occurrence state of the other node is used as the parent node, and the node whose occurrence state is affected is used as the child node. For example, for the existing nodes A and B, if node A affects the occurrence state of node B, then node A is used as the parent node of node B, and node B is used as the child node of node A. Based on the influence relationship between the nodes (parent node and child node), the directed edge information between the nodes (from the parent node to the child node) is determined.

[0055] For example, landslide body characteristics (volume, type of rock and soil) → sliding path; sliding path → debris flow fluid characteristics (capacity, flow velocity, solid particle composition); rainfall intensity → debris flow fluid characteristics; longitudinal gradient of the valley → debris flow velocity; debris flow fluid characteristics → debris flow grade; relative position of the bridge and the channel → bridge damage grade; debris flow grade + bridge structure → bridge damage grade; debris flow grade → bridge damage grade, etc. The arrow direction represents the parent node pointing to the child node.

[0056] Preferably, the calculating sub-module is specifically configured to: discretize the node variables in each historical case, and based on the discretized node variables, use the following formula to calculate the conditional probability between each node to determine the conditional probability table:

[0057]

[0058]

[0059] where l is the number of nodes, P(X1 / X2,X3,...X l ) is the probability when the states of all the parent nodes of the given node V1 are in X2,X3,...X l , and the child node V1 is in the state of X1; P(X1,X2,...X l ) is the joint probability when the states of all the nodes are respectively X1,X2,...X l , and P(Xm = S a ) is the prior probability, P(X i = S b / X m = S a ) is the posterior probability, i.e., the conditional probability, indicating that the states of all parent nodes are S a and the state of the child node is S b The probability of occurrence.

[0060] It can be understood that discretizing the data is a relatively common data preprocessing method, which can effectively simplify the calculation amount and will not be elaborated here.

[0061] Preferably, the model sub-module is specifically used to construct a Bayesian network model BN = (N, E, P). Among them, N is the set of all nodes in the network, E is the set of directed edges connecting each node in the network, that is, the influence relationship (parent node or child node) between each node, and P is the set of conditional probability tables between each node.

[0062] Preferably, the early warning module includes: an acquisition sub-module and an early warning sub-module. Among them, the acquisition sub-module is used to estimate the characteristics of the landslide body, the sliding path and the characteristics of the debris flow fluid corresponding to the slope of the bridge to be measured, and obtain the rainfall intensity, the longitudinal gradient of the gully, the relative position of the bridge and the gully, and the bridge structure corresponding to the slope of the bridge to be measured if the landslide risk level is greater than or equal to the first risk level; the early warning sub-module is used to input the characteristics of the landslide body, the sliding path, the characteristics of the debris flow fluid, the rainfall intensity, the longitudinal gradient of the gully, the relative position of the bridge and the gully, and the bridge structure corresponding to the slope of the bridge to be measured into the early warning model for analysis, obtain the corresponding early warning analysis result, and conduct an early warning based on the early warning analysis result; the early warning analysis result includes the catastrophic risk situation of debris flow and bridge damage.

[0063] In this way, if the landslide risk level is greater than or equal to the first risk level, it means that the probability of a landslide occurring on the bridge slope is relatively high. At this time, the early warning model is used to conduct early warning analysis based on the characteristics of the landslide body, the sliding path, the characteristics of the debris flow fluid, the rainfall intensity, the longitudinal gradient of the gully, the relative position of the bridge and the gully, and the bridge structure corresponding to the slope of the bridge to be measured, so as to be able to conduct disaster early warning before the occurrence of the landslide disaster, and timely and accurately give early warning prompts for the landslide-debris flow-bridge damage disaster chain, reducing disaster losses.

[0064] In some embodiments, the estimation of landslide characteristics, sliding paths, and debris flow fluid characteristics is a relatively common technique and will not be elaborated here. The disaster risk situations specifically include debris flow grade information (each debris flow grade and its corresponding occurrence probability) and bridge damage grade information (each bridge damage grade and its corresponding occurrence probability). The debris flow grade with the highest occurrence probability is taken as the target debris flow grade, and the bridge damage grade with the highest occurrence probability is taken as the target bridge damage grade.

[0065] Preferably, the warning sub-module is specifically configured to: determine a fuzzy warning strategy based on the landslide risk grade and warning analysis result of the landslide occurring on the slope of the bridge to be measured; issue a warning based on the fuzzy warning strategy.

[0066] In this way, when the probability of a landslide occurring on the slope of the bridge to be measured is relatively high, the analysis and warning of the "landslide - debris flow - bridge damage" disaster chain are carried out. Therefore, the probability of a landslide occurring on the slope of the bridge to be measured (i.e., the landslide risk grade) and the warning analysis result are jointly used for warning, and the warning effect is better.

[0067] In some embodiments, the warning sub-module determines the fuzzy warning strategy in the following manner: divide the landslide risk grade into a first target grade and a second target grade, and map them to the membership degrees of a fuzzy set; divide the target debris flow grade into three fuzzy sets: lower, general, and severe, and perform fuzzification through a trapezoidal membership function; divide the target bridge damage grade into three fuzzy sets: minor, general, and severe, and perform fuzzification through a trapezoidal membership function. Construct a fuzzy rule base as follows: 1) When the landslide risk is the second target grade, and (the target debris flow disaster grade, and / or the target bridge damage grade is severe), trigger a no-entry instruction; 2) When the target debris flow grade is severe and / or the target bridge damage grade is severe, regardless of the landslide risk grade, trigger a no-entry instruction; 3) When the target debris flow grade is general, and / or the target bridge damage grade is general), regardless of the landslide risk grade, trigger a restricted-entry instruction; 4) When the landslide risk grade is the first target grade, and the target debris flow grade is lower and the target bridge damage grade is minor, trigger an allow-entry instruction. Based on the three fuzzy sets, perform fuzzy reasoning and defuzzification processing using the fuzzy rule base to determine the output value, and determine the fuzzy warning strategy based on the output value. The fuzzy warning strategy includes: a no-entry instruction, a restricted-entry instruction, and an allow-entry instruction. For example, when the output value ≥ 0.7, generate a no-entry instruction; when 0.4 ≤ output value < 0.7, generate a restricted-entry instruction; when the output value < 0.4, generate an allow-entry instruction.

[0068] Preferably, in combination Figure 2 As shown, the embodiments of the present disclosure provide a method for warning geological disasters on bridge slopes based on multi-source data fusion, including:

[0069] Step S101, obtain the monitoring data of the slope of the bridge to be measured. Among them, the monitoring data includes geological and topographical data, meteorological and hydrological data, engineering activity data, and displacement and deformation data.

[0070] Step S102, determine whether there is abnormal data based on the monitoring data. If there is, use the abnormal data as the preliminary abnormal data.

[0071] Step S103, predict the first probability of landslide occurrence of the slope of the bridge to be measured based on the preliminary abnormal data, and predict the second probability of landslide occurrence of the slope of the bridge to be measured based on the monitoring data, and determine the landslide risk level of the slope of the bridge to be measured based on the first probability and the second probability.

[0072] Step S104, if the landslide risk level is greater than or equal to the first risk level, use the preset early warning model to conduct early warning analysis of the disaster chain of landslide-debris flow-bridge damage, obtain the early warning analysis result, and conduct early warning based on the early warning analysis result.

[0073] Adopt the bridge slope geological disaster early warning method based on multi-source data fusion provided by the embodiments of the present disclosure. By obtaining the monitoring data of the slope of the bridge to be measured, the abnormal data in the monitoring data is used as the preliminary abnormal data. The preliminary abnormal data is used to reflect the local abnormality of the slope of the bridge to be measured, while the monitoring data reflects the overall state of the bridge slope. In this way, the first probability of landslide occurrence of the slope of the bridge to be measured is predicted using the preliminary data, the second probability of landslide occurrence of the slope of the bridge to be measured is predicted using the monitoring data, and then the landslide risk level of the slope of the bridge to be measured is determined based on the first probability and the second probability, so as to realize the prediction of whether the slope of the bridge to be measured will have a landslide based on the local abnormality and the overall state of the slope of the bridge to be measured, so as to be able to more accurately predict whether the bridge slope will have a landslide disaster. When the landslide risk level is greater than or equal to the first risk level, it means that the probability of landslide occurrence of the slope of the bridge to be measured is relatively large at this time. At this time, early warning analysis of the disaster chain of landslide-debris flow-bridge damage is carried out, so as to be able to timely and accurately give early warning prompts for the disaster chain of landslide-debris flow-bridge damage.

[0074] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made in form and details without departing from the scope defined by the claims of the present invention.

Claims

1. A bridge slope geological disaster warning system based on multi-source data fusion, characterized in that, Including: An acquisition module, configured to acquire monitoring data of a bridge slope to be measured; where the monitoring data includes geological and topographical data, meteorological and hydrological data, engineering activity data, and displacement and deformation data; A processing module, configured to determine whether there is abnormal data based on the monitoring data, and if so, use the abnormal data as preliminary abnormal data; A prediction module, configured to predict a first probability of landslide occurrence of the bridge slope to be measured based on the preliminary abnormal data, and predict a second probability of landslide occurrence of the bridge slope to be measured based on the monitoring data, and determine a landslide risk level of the bridge slope to be measured based on the first probability and the second probability; An early warning module, configured to, if the landslide risk level is greater than or equal to a first risk level, perform disaster chain early warning analysis of landslide-debris flow-bridge damage using a preset early warning model, obtain an early warning analysis result, and perform early warning based on the early warning analysis result.

2. The system according to claim 1, wherein The prediction module includes: A first sub-module, configured to calculate a sub-probability of each piece of preliminary abnormal data causing a landslide of the bridge slope to be measured based on each piece of preliminary abnormal data; A second sub-module, configured to comprehensively determine the first probability based on each sub-probability.

3. The system according to claim 2, wherein The first sub-module includes: A first calculation unit, configured to calculate the probability of natural instability of the bridge slope to be measured, calculate the correlation degree between each piece of preliminary abnormal data and the bridge slope to be measured, and the probability of occurrence of each piece of preliminary abnormal data; A second calculation unit, configured to, for each piece of preliminary abnormal data, calculate the sub-probability of each piece of preliminary abnormal data causing a landslide of the bridge slope to be measured using Bayes' formula based on the probability of natural instability, the correlation degree, and the probability of occurrence of the preliminary abnormal data, as follows: Among them, P(A / B i ) is the sub-probability that the landslide of the bridge slope to be measured is caused by the i-th type of prior abnormal data, P(B i / A) is the correlation degree between the i-th type of prior abnormal data and the bridge slope to be measured, P(A) is the instability probability, and P(B i ) is the probability that the i-th type of prior abnormal data appears.

4. The system according to claim 3, characterized in that, The first calculation unit is specifically configured to: Obtain the total number of slopes in a preset monitoring area and the number of slopes that have undergone natural instability; Based on the total number of slopes in the preset monitoring area and the number of slopes that have undergone natural instability in the preset monitoring area, use the classical probability formula to calculate the probability of natural instability of the slopes in the preset monitoring area, as follows: Where P(C) is the probability of natural instability of the slopes in the preset monitoring area, m is the total number of slopes in the preset monitoring area, and n is the number of slopes that have undergone natural instability in the preset monitoring area; Use the probability of natural instability of the slopes in the preset monitoring area as the probability of natural instability of the bridge slope to be measured.

5. The system according to claim 1, wherein The prediction module includes a third sub-module; the third sub-module is specifically configured to: Respectively determine a first level corresponding to the first probability and a second level corresponding to the second probability; If any one of the first level or the second level is greater than or equal to a second risk level, determine the highest level among the first level and the second level as the landslide risk level of the bridge slope to be measured; If the first level and the second level are both lower than the second risk level, the landslide risk level of the bridge slope to be tested is determined based on the first level, the first preset weight corresponding to the first level, the second level and the second preset weight corresponding to the second level; wherein the second risk level is greater than the first risk level.

6. The system according to any one of claims 1 to 5, characterized in that, Also included is a model building module; the model building module includes: The node submodule is used to obtain nodes related to the "landslide-mud flow-bridge damage" disaster chain; the nodes include three types of nodes: disaster-causing factors, disaster-pregnant environments, and disaster events. The disaster-causing factor nodes include landslide body characteristics and sliding paths, the disaster-pregnant environment nodes include rainfall intensity, valley longitudinal gradient, mud flow fluid characteristics, bridge and ditch relative position and bridge structure, and the disaster event nodes include mud flow grade and bridge damage grade. A determination submodule is used to determine the influence relationship between each node based on the disaster mechanism and chain transmission law of the "landslide-mudslide-bridge damage" disaster chain, and determine the directed edge information between each node based on the influence relationship; A case submodule, for obtaining corresponding multiple historical cases based on the "landslide-mudslide-bridge damage" disaster chain; the historical cases include node variables corresponding to each node; The calculation submodule is used to calculate the conditional probability between each node using the node variables of multiple historical cases; The model submodule is used to construct a Bayesian network model based on each node, directed edge information between each node, and conditional probability between each node, and determine the constructed Bayesian network model as an early warning model.

7. The system according to claim 6, wherein The early warning module comprises: An acquisition submodule is used for estimating the landslide body characteristics, sliding path and debris flow fluid characteristics corresponding to the bridge slope to be tested, and obtaining the rainfall intensity, valley longitudinal gradient, relative position of the bridge and the channel and bridge structure corresponding to the bridge slope to be tested if the landslide risk level is greater than or equal to the first risk level; The early warning submodule is used to input the landslide characteristics, sliding path, debris flow fluid characteristics, rainfall intensity, valley longitudinal gradient, relative position of the bridge and the ditch, and bridge structure corresponding to the bridge slope to be tested into the early warning model for analysis, obtain corresponding early warning analysis results, and issue an early warning based on the early warning analysis results; the early warning analysis results include the catastrophic risk of debris flow and bridge damage.

8. The system according to claim 7, characterized in that, The early warning submodule is specifically used for: Determine a fuzzy early warning strategy based on the landslide risk level of the bridge slope to be tested and the early warning analysis result; An early warning is performed based on the fuzzy early warning strategy.

9. A method for warning of geological disasters on bridge slopes based on multi-source data fusion, characterized in that, include: Obtain monitoring data of the bridge slope to be tested; wherein the monitoring data includes geological and topographic data, meteorological and hydrological data, engineering activity data and displacement and deformation data; Determine whether there is abnormal data based on the monitoring data, and if so, take the abnormal data as prior abnormal data; Predict the first probability of landslide occurring on the slope of the bridge to be measured based on the prior abnormal data, and predict the second probability of landslide occurring on the slope of the bridge to be measured based on the monitoring data, and determine the landslide risk level of the slope of the bridge to be measured based on the first probability and the second probability; If the landslide risk level is greater than or equal to the first risk level, use a preset early warning model to conduct a disaster chain early warning analysis of landslide-debris flow-bridge damage, obtain the early warning analysis result, and conduct an early warning based on the early warning analysis result.

Images