Debris flow monitoring and early warning method and system

Abstract

The invention discloses a debris flow monitoring and early warning method and system. The method comprises the following steps: constructing a debris flow susceptible area judgment model, and calculating a debris flow susceptible area judgment value based on rainfall prediction data, debris flow distribution historical data, human production data and the debris flow susceptible area judgment model so as to determine a key area of the debris flow susceptible area; arranging a sensor network, and obtaining multi-index sensor data of a monitoring area; obtaining a deformation value of the monitoring area based on the high-resolution satellite remote sensing image data of the monitoring area; and constructing a prediction model of the debris flow in the monitoring area, and obtaining a prediction result of the debris flow in the monitoring area so as to execute a corresponding early warning operation on the monitoring area. According to the invention, the monitoring range of the debris flow area can be narrowed, the debris flow monitoring cost is reduced, the deformation value of the monitoring area and the multi-index sensor data are fused, the accuracy of the prediction result can be effectively improved, and the early warning operation is more practical.

Description

Technical Field

[0001] The present invention relates to the field of debris flow monitoring and early warning, and in particular to a debris flow monitoring and early warning method and system. Background Art

[0002] Debris flow is a sudden natural disaster, and its occurrence time and place are highly uncertain. Once a debris flow occurs, its destructive power is extremely large, often causing great harm to human life and property. Therefore, the prediction of debris flow is particularly important, which can provide scientific basis and technical support for debris flow warning and prevention.

[0003] At present, the debris flow warning method can predict debris flows to a certain extent by combining machine learning methods and satellite remote sensing methods. However, the existing warning method can only cover specific areas, and the impact range of debris flows may exceed the coverage of the warning system, resulting in some areas being unable to receive warning information in time. In addition, due to errors in monitoring equipment or limitations of algorithms, the existing warning methods will issue erroneous warning information, causing people to take unnecessary evacuation actions, wasting resources and energy. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a debris flow monitoring and early warning method and system.

[0005] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is:

[0006] A debris flow monitoring and early warning method comprises the following steps:

[0007] S1. Construct a debris flow prone area determination model, calculate the debris flow prone area determination value based on rainfall forecast data, debris flow distribution historical data, human production data and the debris flow prone area determination model, and determine the key areas of the debris flow prone area according to the debris flow prone area determination value;

[0008] S2, deploying a sensor network based on the key areas of the debris flow prone area in step S1, and acquiring multi-index sensor data of the monitoring area;

[0009] S3, based on the key area of ​​the debris flow prone area in step S1, obtain high-resolution satellite remote sensing image data of the monitoring area, and based on the high-resolution satellite remote sensing image data of the monitoring area, use the least squares method and singular value decomposition method to obtain the deformation value of the monitoring area;

[0010] S4. Construct a prediction model for debris flow in the monitoring area, obtain the prediction result of debris flow in the monitoring area based on the prediction model for debris flow in the monitoring area, the multi-index sensor data of the monitoring area in step S2 and the deformation value of the monitoring area in step S3, and perform early warning operations on the monitoring area based on the prediction result of debris flow in the monitoring area.

[0011] Furthermore, in step S1, a debris flow prone area determination model is constructed, which is expressed as:

[0012]

[0013] Among them: K is the judgment value of the debris flow prone area in the region, α is the weight of the predicted rainfall balance index, R is the predicted rainfall balance index value of the region, i is the sequence number of the human production index, n is the number of human production indicators, β i is the weight of the i-th human production index, V i Produce an indicator value for the i-th human.

[0014] Further, in step S1, a debris flow prone area determination value is calculated based on rainfall forecast data, debris flow distribution historical data, human production data and a debris flow prone area determination model, and a key area of ​​the debris flow prone area is determined according to the debris flow prone area determination value, including the following steps:

[0015] A1. Initialize the weight of the predicted rainfall balance index and the weight of the human production index, calculate the rainfall balance index value of the region based on the rainfall prediction data, and determine the human production index value based on the human production data;

[0016] A2, using the weight of the predicted rainfall balance index, the weight of the human production index, the rainfall balance index value of the region, the human production index value and the debris flow prone area determination model in step A1 to obtain the debris flow prone area determination training value of the region;

[0017] A3, obtaining a debris flow determination training result for the region according to the debris flow prone region determination threshold and the debris flow prone region determination training value in step A2;

[0018] A4. Calibrate the debris flow determination training results of the area in step A3 based on the debris flow distribution historical data to update the debris flow prone area determination model, calculate the debris flow prone area determination value using the updated debris flow prone area determination model, and determine the key area of ​​the debris flow prone area according to the debris flow prone area determination value.

[0019] Furthermore, in step S2, a sensor network is arranged based on the key areas of the debris flow-prone areas in step S1. The specific process is: based on the key areas of the debris flow-prone areas in step S1, a rain gauge is set at the valley entrance in the key area, a soil moisture meter is set at the turning point in the key area, and a displacement sensor is set in the accumulation area of ​​the key area to complete the arrangement of the sensor network.

[0020] Further, in step S3, the deformation value of the monitoring area is obtained based on the high-resolution satellite remote sensing image data of the monitoring area, including the following steps:

[0021] B1. Determine a main image based on the high-resolution satellite remote sensing image data of the monitoring area, register the remaining auxiliary images, and obtain the registered high-resolution satellite remote sensing image data;

[0022] B2, combining the high-resolution satellite remote sensing image data registered in step B1 in pairs to generate differential interferometry pairs, applying the least square method to the differential interferometry pairs to obtain the deformation within the short baseline set, and applying the singular value decomposition method to the deformation within the short baseline set to obtain the deformation information of the entire baseline set;

[0023] B3. Remove the flat ground phase and terrain phase from the deformation information of the entire baseline set in step B2, perform phase unwrapping on the differential interferometer pair from which the flat ground phase and terrain phase are removed according to the coherence coefficient diagram to obtain the unwrapped differential interferometer phase, and perform two model inversions and geocoding of the differential interferometer measurement short baseline set time series analysis processing on the unwrapped differential interferometer phase to obtain the deformation value of the monitoring area.

[0024] Further, in step B2, the deformation phase of the differential interference pair is expressed as:

[0025]

[0026] in: is the deformation phase of the jth differential interferometer pair, j is the serial number of the differential interferometer pair, A is the time period corresponding to the acquisition of the primary image of the jth differential interferometer pair, B is the time period corresponding to the acquisition of the auxiliary image of the jth differential interferometer pair, k is the serial number of the time period, t B，j+1 is the time corresponding to the image of the j+1th differential interference pair in time period B, t A，j is the time corresponding to the image of the jth differential interference pair in time period A, t k is the kth moment, t k-1 is the k-1th moment, v k is the average deformation rate at the kth moment in the image sequence.

[0027] Further, in step S4, the constructed prediction model for debris flow in the monitoring area includes a first hidden layer, a second hidden layer and a third linear layer connected in sequence;

[0028] The first hidden layer is used to receive the monitoring area input data, use the first weight term coefficient and the first bias term coefficient to increase the dimension of the monitoring area input data to obtain the first high-dimensional monitoring area input data, use the maximum function to perform a nonlinear transformation on the first high-dimensional monitoring area input data to obtain the first high-dimensional enhanced monitoring area input data, and use the regularization method to reduce the dimension of the first high-dimensional enhanced monitoring area input data to obtain the second monitoring area input data;

[0029] The second hidden layer is used to reduce the dimension of the second monitoring area input data according to the second weight item coefficient and the second bias item coefficient to obtain the second low-dimensional monitoring area input data, use the maximum function to perform a nonlinear transformation on the second low-dimensional monitoring area input data to obtain the second low-dimensional enhanced monitoring area input data, and use the regularization method to reduce the dimension of the second low-dimensional enhanced monitoring area input data to obtain the third monitoring area input data.

[0030] The third linear layer is used to input data of the third monitoring area into the three-dimensional space according to the third weight term coefficient and the third bias term coefficient to obtain the prediction result of the debris flow occurring in the monitoring area.

[0031] Furthermore, the data processing process of the first hidden layer is expressed as:

[0032] X2=max(X1W1+b1,0)⊙M

[0033] Where: X2 is the input data of the second monitoring area, max is the maximum value symbol, X1 is the input data of the monitoring area, X1∈R N×F , N is the number of samples of input data in the monitoring area, F is the characteristic dimension of pollution source data, F=65, W1 is the first weight term coefficient, b1 is the first bias term coefficient, ⊙ is element multiplication, M is the mask matrix, and the elements in the mask matrix are 0 or 1.

[0034] A debris flow monitoring and early warning system using the above method comprises a monitoring area determination unit, a data acquisition unit, a deformation calculation unit and a monitoring and early warning unit;

[0035] The monitoring area determination unit is used to construct a debris flow prone area determination model, calculate the debris flow prone area determination value based on rainfall forecast data, debris flow distribution historical data, human production data and the debris flow prone area determination model, and determine the key areas of the debris flow prone area according to the debris flow prone area determination value;

[0036] The data acquisition unit is used to arrange the sensor network according to the key areas of the debris flow prone area and obtain the multi-index sensor data of the monitoring area;

[0037] The deformation calculation unit is used to obtain high-resolution satellite remote sensing image data of the monitoring area according to the key areas of the debris flow prone area, and based on the high-resolution satellite remote sensing image data of the monitoring area, the least square method and singular value decomposition method are used to obtain the deformation value of the monitoring area;

[0038] The monitoring and early warning unit is used to construct a prediction model for debris flow in the monitoring area, obtain the prediction results of debris flow in the monitoring area based on the prediction model for debris flow in the monitoring area, the multi-indicator sensor data of the monitoring area and the deformation value of the monitoring area, and perform early warning operations on the monitoring area based on the prediction results of debris flow in the monitoring area.

[0039] The present invention has the following beneficial effects:

[0040] (1) The present invention constructs a debris flow prone area determination model, calculates a debris flow prone area determination value based on rainfall forecast data, debris flow distribution historical data, human production data and the debris flow prone area determination model, and determines the key area of ​​the debris flow prone area according to the debris flow prone area determination value, thereby narrowing the monitoring scope of the debris flow area, improving the accuracy of debris flow monitoring from a global perspective, and reducing the cost of debris flow monitoring;

[0041] (2) The present invention is based on the key areas of debris flow prone areas, and sets rain gauges at the valley entrances in the key areas, sets soil moisture meters at the turning points in the key areas, and sets displacement sensors in the accumulation areas in the key areas, which can improve the accuracy of multi-index sensor data in the monitoring area;

[0042] (3) The present invention constructs a prediction model for debris flow in the monitoring area, including a first hidden layer, a second hidden layer and a third linear layer connected in sequence, so that the parameter quantity and structural complexity of the prediction model for debris flow in the monitoring area can match the data quantity and complexity of the input data in the monitoring area, thereby avoiding the occurrence of overfitting while ensuring the accuracy of subsequent prediction results:

[0043] (4) The present invention proposes a debris flow monitoring and early warning system, comprising a monitoring area determination unit, a data acquisition unit, a deformation calculation unit and a monitoring and early warning unit; the monitoring area determination unit can be used to construct a debris flow prone area determination model, calculate a debris flow prone area determination value based on rainfall forecast data, debris flow distribution historical data, human production data and the debris flow prone area determination model, and determine the key area of ​​the debris flow prone area according to the debris flow prone area determination value; the data acquisition unit can be used to arrange a sensor network according to the key area of ​​the debris flow prone area, and obtain multi-index sensor data of the monitoring area; The deformation calculation unit can be used to obtain high-resolution satellite remote sensing image data of the monitoring area according to the key areas of the debris flow prone area, and based on the high-resolution satellite remote sensing image data of the monitoring area, the least squares method and singular value decomposition method are used to obtain the deformation value of the monitoring area; the monitoring and early warning unit can be used to construct a prediction model for the occurrence of debris flow in the monitoring area, and obtain the prediction results of the occurrence of debris flow in the monitoring area based on the prediction model for the occurrence of debris flow in the monitoring area, the multi-index sensor data of the monitoring area and the deformation value of the monitoring area, and perform early warning operations on the monitoring area based on the prediction results of the occurrence of debris flow in the monitoring area. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a schematic diagram of a debris flow monitoring and early warning method;

[0045] Figure 2 The figure is a schematic diagram of the structure of a debris flow monitoring and early warning system. DETAILED DESCRIPTION

[0046] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0047] like Figure 1 As shown, a debris flow monitoring and early warning method includes steps S1-S4, which are as follows:

[0048] S1. Construct a debris flow prone area determination model, calculate the debris flow prone area determination value based on rainfall forecast data, debris flow distribution historical data, human production data and the debris flow prone area determination model, and determine the key areas of the debris flow prone area according to the debris flow prone area determination value.

[0049] In an optional embodiment of the present invention, the present invention constructs a debris flow prone area determination model, which is expressed as:

[0050]

[0051] Among them: K is the judgment value of the debris flow prone area in the region, α is the weight of the predicted rainfall balance index, R is the predicted rainfall balance index value of the region, i is the sequence number of the human production index, n is the number of human production indicators, β i is the weight of the i-th human production index, V i Produce an indicator value for the i-th human.

[0052] The present invention calculates a debris flow prone area determination value based on rainfall forecast data, debris flow distribution historical data, human production data and a debris flow prone area determination model, and determines a key area of ​​the debris flow prone area according to the debris flow prone area determination value, comprising the following steps:

[0053] A1. Initialize the weight of the predicted rainfall balance index and the weight of the human production index, calculate the rainfall balance index value of the region based on the rainfall prediction data, and determine the human production index value based on the human production data.

[0054] The present invention calculates a rainfall balance index value of a region based on rainfall prediction data, and the specific process is: based on the rainfall prediction data, the predicted cumulative rainfall of the region in the next month is divided by 30 to obtain a first prediction value, the predicted cumulative rainfall of the region in the next week is divided by 7 to obtain a second prediction value, the predicted cumulative rainfall of the region in the next three days is divided by 3 to obtain a third prediction value, the predicted cumulative rainfall of the region in the next twenty-four hours is determined as a fourth prediction value, and the average of the first prediction value, the second prediction value, the third prediction value and the fourth prediction value is calculated to obtain the average of the prediction values ​​of the region, and the average daily rainfall in the same period in history is subtracted from the average of the prediction values ​​of the region to obtain the rainfall balance index value of the region.

[0055] Human production indicators include population density indicators, housing construction quantity indicators and road and bridge quantity indicators.

[0056] A2. Obtain the training value for determining the debris flow prone area in the region by using the weight of the rainfall balance index predicted in step A1, the weight of the human production index, the regional rainfall balance index value, the human production index value and the debris flow prone area determination model.

[0057] A3. Obtaining a debris flow determination training result for a region according to the debris flow prone region determination threshold and the debris flow prone region determination training value in step A2.

[0058] The present invention determines whether the mudslide prone area determination training value of a region is greater than the mudslide prone area determination threshold; if so, the mudslide determination training result of the region is determined as a key area of ​​the mudslide prone area, otherwise it is not determined as the key area of ​​the mudslide prone area.

[0059] A4. Calibrate the debris flow determination training results of the area in step A3 based on the debris flow distribution historical data to update the debris flow prone area determination model, calculate the debris flow prone area determination value using the updated debris flow prone area determination model, and determine the key area of ​​the debris flow prone area according to the debris flow prone area determination value.

[0060] The present invention judges whether the mudslide determination training result of a region is accurate based on the mudslide distribution historical data. If so, the weight of the predicted rainfall balance index and the weight of the human production index are adjusted according to the asymptotic principle, and the mudslide prone area determination model after adjusting the weight of the predicted rainfall balance index and the weight of the human production index is used to obtain the mudslide determination training result of the region, and then continues to judge whether the mudslide determination training result is accurate, until the accuracy rate of the mudslide determination training result reaches 90%, the update is completed, and the mudslide prone area determination model at this time is determined as the updated mudslide prone area determination model.

[0061] S2. Arrange a sensor network based on the key areas of the debris flow prone area in step S1 to obtain multi-index sensor data in the monitoring area.

[0062] In an optional embodiment of the present invention, the present invention arranges a sensor network based on the key areas of the debris flow prone areas in step S1. The specific process is: based on the key areas of the debris flow prone areas in step S1, a rain gauge is set at the entrance of the valley in the key area, a soil moisture meter is set at the turning point of the key area, and a displacement sensor is set in the accumulation area of ​​the key area to complete the arrangement of the sensor network.

[0063] S3. Based on the key areas of the debris flow-prone areas in step S1, high-resolution satellite remote sensing image data of the monitoring area are obtained. Based on the high-resolution satellite remote sensing image data of the monitoring area, the least squares method and singular value decomposition method are used to obtain the deformation value of the monitoring area.

[0064] In an optional embodiment of the present invention, the present invention obtains the deformation value of the monitoring area based on the high-resolution satellite remote sensing image data of the monitoring area, including the following steps:

[0065] B1. Determine a main image based on the high-resolution satellite remote sensing image data of the monitoring area, align the remaining auxiliary images, and obtain the aligned high-resolution satellite remote sensing image data.

[0066] B2. The high-resolution satellite remote sensing image data registered in step B1 are combined in pairs to generate differential interferometry pairs, and the least squares method is used for the differential interferometry pairs to obtain the deformation within the short baseline set, and the singular value decomposition method is used for the deformation within the short baseline set to obtain the deformation information of the entire baseline set.

[0067] Specifically, the present invention sets corresponding time and space baseline thresholds for the registered high-resolution satellite remote sensing image data, and combines the registered high-resolution satellite remote sensing image data that meet the space baseline thresholds in the same time in pairs to generate differential interference pairs.

[0068] The deformation phase of the differential interference pair is expressed as:

[0069]

[0070] in: is the deformation phase of the jth differential interferometer pair, j is the serial number of the differential interferometer pair, A is the time period corresponding to the acquisition of the primary image of the jth differential interferometer pair, B is the time period corresponding to the acquisition of the auxiliary image of the jth differential interferometer pair, k is the serial number of the time period, t B，j+1 is the time corresponding to the image of the j+1th differential interference pair in time period B, t A，j is the time corresponding to the image of the jth differential interference pair in time period A, t k is the kth moment, t k-1 is the k-1th moment, v k is the average deformation rate at the kth moment in the image sequence.

[0071] B3. Remove the flat ground phase and terrain phase from the deformation information of the entire baseline set in step B2, perform phase unwrapping on the differential interferometer pair from which the flat ground phase and terrain phase are removed according to the coherence coefficient diagram to obtain the unwrapped differential interferometer phase, and perform two model inversions and geocoding of the differential interferometer measurement short baseline set time series analysis processing on the unwrapped differential interferometer phase to obtain the deformation value of the monitoring area.

[0072] The present invention combines satellite orbit data, imaging geometry model and external digital elevation model data to remove the flat ground phase and terrain phase from the deformation information of the entire baseline set. The present invention performs phase unwrapping on the interference graph from which the flat ground phase and terrain phase are removed according to the coherence coefficient graph, and can obtain the unwrapped differential interference phase without digital elevation model error.

[0073] S4. Construct a prediction model for debris flow in the monitoring area, obtain the prediction result of debris flow in the monitoring area based on the prediction model for debris flow in the monitoring area, the multi-index sensor data of the monitoring area in step S2 and the deformation value of the monitoring area in step S3, and perform early warning operations on the monitoring area based on the prediction result of debris flow in the monitoring area.

[0074] In an optional embodiment of the present invention, the present invention standardizes the multi-index sensor data of the monitoring area and the deformation value of the monitoring area to obtain the input data of the monitoring area, and uses the input data of the monitoring area to train, verify and test the prediction model of the occurrence of debris flow in the monitoring area, wherein the cross entropy loss function is used to iterate the prediction model of the occurrence of debris flow in the monitoring area to obtain the predicted model of the occurrence of debris flow in the monitoring area after the test, and the predicted result of the occurrence of debris flow in the monitoring area is obtained by using the predicted model of the occurrence of debris flow in the monitoring area after the test and the latest input data of the monitoring area.

[0075] The prediction model for debris flow in the monitoring area constructed by the present invention comprises a first hidden layer, a second hidden layer and a third linear layer connected in sequence. The present invention adopts the first hidden layer and the second hidden layer, so that the parameter quantity and structural complexity of the prediction model for debris flow in the monitoring area can match the data quantity and complexity of the input data in the monitoring area, and avoids the occurrence of overfitting under the premise of ensuring the accuracy of subsequent prediction results.

[0076] The first hidden layer is used to receive the input data of the monitoring area, use the first weight item coefficient and the first bias item coefficient to increase the dimension of the input data of the monitoring area to obtain the first high-dimensional monitoring area input data, use the maximum function to perform a nonlinear transformation on the first high-dimensional monitoring area input data to obtain the first high-dimensional enhanced monitoring area input data, and use the regularization method to reduce the dimension of the first high-dimensional enhanced monitoring area input data to obtain the second monitoring area input data.

[0077] The data processing process of the first hidden layer is expressed as:

[0078] X2=max(X1W1+b1,0)⊙M

[0079] Where: X2 is the input data of the second monitoring area, max is the maximum value symbol, X1 is the input data of the monitoring area, X1∈R N×F , N is the number of samples of input data in the monitoring area, F is the characteristic dimension of pollution source data, F=65, W1 is the first weight term coefficient, b1 is the first bias term coefficient, ⊙ is element multiplication, M is the mask matrix, and the elements in the mask matrix are 0 or 1.

[0080] The second hidden layer is used to reduce the dimension of the second monitoring area input data according to the second weight item coefficient and the second bias item coefficient to obtain the second low-dimensional monitoring area input data, use the maximum function to perform a nonlinear transformation on the second low-dimensional monitoring area input data to obtain the second low-dimensional enhanced monitoring area input data, and use the regularization method to reduce the dimension of the second low-dimensional enhanced monitoring area input data to obtain the third monitoring area input data.

[0081] The data processing process of the second hidden layer is expressed as:

[0082] X3=max(X2W2+b2,0)⊙M

[0083] Wherein: X3 is the input data of the third monitoring area, W2 is the second weight term coefficient, and b2 is the second bias term coefficient.

[0084] The third linear layer is used to input data of the third monitoring area into the three-dimensional space according to the third weight term coefficient and the third bias term coefficient to obtain the prediction result of the debris flow occurring in the monitoring area.

[0085] The data processing process of the third linear layer is expressed as:

[0086] X=X3W3+b3

[0087] Where: X is the prediction result of debris flow in the monitoring area, W3 is the third weight term coefficient, and b3 is the third bias term coefficient.

[0088] The present invention performs early warning operations on the monitoring area based on the prediction result of mudslides in the monitoring area: if the prediction result of mudslides in the monitoring area is no, no early warning warning is performed; if the prediction result of mudslides in the monitoring area is yes, early warning warning is performed through the LED indicator light.

[0089] like Figure 2 As shown, a debris flow monitoring and early warning system using the above method includes a monitoring area determination unit, a data acquisition unit, a deformation calculation unit and a monitoring and early warning unit.

[0090] In an optional embodiment of the present invention, the monitoring area determination unit is used to construct a debris flow prone area determination model, calculate the debris flow prone area determination value based on rainfall forecast data, debris flow distribution historical data, human production data and the debris flow prone area determination model, and determine the key areas of the debris flow prone area according to the debris flow prone area determination value.

[0091] In an optional embodiment of the present invention, the data acquisition unit is used to arrange a sensor network according to key areas of the debris flow prone area to obtain multi-index sensor data of the monitoring area.

[0092] In an optional embodiment of the present invention, the deformation calculation unit is used to obtain high-resolution satellite remote sensing image data of the monitoring area based on the key areas of the debris flow prone area, and based on the high-resolution satellite remote sensing image data of the monitoring area, the least squares method and singular value decomposition method are used to obtain the deformation value of the monitoring area.

[0093] In an optional embodiment of the present invention, the monitoring and early warning unit is used to construct a prediction model for the occurrence of debris flows in the monitoring area, obtain the prediction results of the occurrence of debris flows in the monitoring area based on the prediction model for the occurrence of debris flows in the monitoring area, the multi-indicator sensor data of the monitoring area and the deformation value of the monitoring area, and perform early warning operations on the monitoring area based on the prediction results of the occurrence of debris flows in the monitoring area.

[0094] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0095] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0096] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0097] The present invention uses specific embodiments to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

[0098] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A debris flow monitoring and early warning method, characterized in that: The following steps are involved: S1. Construct a debris flow prone area determination model, calculate the debris flow prone area determination value based on rainfall forecast data, debris flow distribution historical data, human production data and the debris flow prone area determination model, and determine the key areas of the debris flow prone area according to the debris flow prone area determination value; S2, deploying a sensor network based on the key areas of the debris flow prone area in step S1, and acquiring multi-index sensor data of the monitoring area; S3, based on the key area of ​​the debris flow prone area in step S1, obtain high-resolution satellite remote sensing image data of the monitoring area, and based on the high-resolution satellite remote sensing image data of the monitoring area, use the least squares method and singular value decomposition method to obtain the deformation value of the monitoring area; S4. Construct a prediction model for debris flow in the monitoring area, obtain the prediction result of debris flow in the monitoring area based on the prediction model for debris flow in the monitoring area, the multi-index sensor data of the monitoring area in step S2 and the deformation value of the monitoring area in step S3, and perform early warning operations on the monitoring area based on the prediction result of debris flow in the monitoring area.

2. The debris flow monitoring and early warning method according to claim 1 is characterized in that: In step S1, a debris flow prone area determination model is constructed, which is expressed as: Among them: K is the judgment value of the debris flow prone area in the region, α is the weight of the predicted rainfall balance index, R is the predicted rainfall balance index value of the region, i is the sequence number of the human production index, n is the number of human production indicators, β i is the weight of the i-th human production indicator, t i Produce an indicator value for the i-th human.

3. The debris flow monitoring and early warning method according to claim 1 is characterized in that: In step S1, a debris flow prone area determination value is calculated based on rainfall forecast data, debris flow distribution historical data, human production data and a debris flow prone area determination model, and a key area of ​​the debris flow prone area is determined according to the debris flow prone area determination value, including the following steps: A1. Initialize the weight of the predicted rainfall balance index and the weight of the human production index, calculate the rainfall balance index value of the region based on the rainfall prediction data, and determine the human production index value based on the human production data; A2, using the weight of the predicted rainfall balance index, the weight of the human production index, the rainfall balance index value of the region, the human production index value and the debris flow prone area determination model in step A1 to obtain the debris flow prone area determination training value of the region; A3, obtaining a debris flow determination training result for the region according to the debris flow prone region determination threshold and the debris flow prone region determination training value in step A2; A4. Calibrate the debris flow determination training results of the area in step A3 based on the debris flow distribution historical data to update the debris flow prone area determination model, calculate the debris flow prone area determination value using the updated debris flow prone area determination model, and determine the key area of ​​the debris flow prone area according to the debris flow prone area determination value.

4. The debris flow monitoring and early warning method according to claim 1 is characterized in that: In step S2, a sensor network is arranged based on the key areas of the debris flow prone areas in step S1. The specific process is: based on the key areas of the debris flow prone areas in step S1, a rain gauge is set at the valley entrance in the key area, a soil moisture meter is set at the turning point in the key area, and a displacement sensor is set in the accumulation area of ​​the key area to complete the arrangement of the sensor network.

5. The debris flow monitoring and early warning method according to claim 1 is characterized in that: In step S3, the deformation value of the monitoring area is obtained based on the high-resolution satellite remote sensing image data of the monitoring area, including the following steps: B1. Determine a main image based on the high-resolution satellite remote sensing image data of the monitoring area, register the remaining auxiliary images, and obtain the registered high-resolution satellite remote sensing image data; B2, combining the high-resolution satellite remote sensing image data registered in step B1 in pairs to generate differential interferometry pairs, applying the least square method to the differential interferometry pairs to obtain the deformation within the short baseline set, and applying the singular value decomposition method to the deformation within the short baseline set to obtain the deformation information of the entire baseline set; B3. Remove the flat ground phase and terrain phase from the deformation information of the entire baseline set in step B2, perform phase unwrapping on the differential interferometer pair from which the flat ground phase and terrain phase are removed according to the coherence coefficient diagram to obtain the unwrapped differential interferometer phase, and perform two model inversions and geocoding of the differential interferometer measurement short baseline set time series analysis processing on the unwrapped differential interferometer phase to obtain the deformation value of the monitoring area.

6. The debris flow monitoring and early warning method according to claim 5 is characterized in that: In step B2, the deformation phase of the differential interference pair is expressed as: in: is the deformation phase of the jth differential interferometer pair, j is the serial number of the differential interferometer pair, A is the time period corresponding to the acquisition of the primary image of the jth differential interferometer pair, B is the time period corresponding to the acquisition of the auxiliary image of the jth differential interferometer pair, k is the serial number of the time period, t B，j+1 is the time corresponding to the image of the j+1th differential interference pair in time period B, t A，j is the time t corresponding to the image of the jth differential interference pair in time period A k is the kth moment, t k-1 is the k-1th moment, v k is the average deformation rate at the kth moment in the image sequence.

7. The debris flow monitoring and early warning method according to claim 1 is characterized in that: In step S4, the constructed prediction model for debris flow in the monitoring area includes a first hidden layer, a second hidden layer and a third linear layer connected in sequence; The first hidden layer is used to receive the monitoring area input data, use the first weight term coefficient and the first bias term coefficient to increase the dimension of the monitoring area input data to obtain the first high-dimensional monitoring area input data, use the maximum function to perform a nonlinear transformation on the first high-dimensional monitoring area input data to obtain the first high-dimensional enhanced monitoring area input data, and use the regularization method to reduce the dimension of the first high-dimensional enhanced monitoring area input data to obtain the second monitoring area input data; The second hidden layer is used to reduce the dimension of the second monitoring area input data according to the second weight item coefficient and the second bias item coefficient to obtain the second low-dimensional monitoring area input data, use the maximum function to perform a nonlinear transformation on the second low-dimensional monitoring area input data to obtain the second low-dimensional enhanced monitoring area input data, and use the regularization method to reduce the dimension of the second low-dimensional enhanced monitoring area input data to obtain the third monitoring area input data. The third linear layer is used to input data of the third monitoring area into the three-dimensional space according to the third weight term coefficient and the third bias term coefficient to obtain the prediction result of the debris flow occurring in the monitoring area.

8. The debris flow monitoring and early warning method according to claim 7 is characterized in that: The data processing process of the first hidden layer is expressed as: X2=max(X1W1+b1,0)⊙M Where: X2 is the input data of the second monitoring area, max is the maximum value symbol, X1 is the input data of the monitoring area, X1∈R N×F , N is the number of samples of input data in the monitoring area, F is the characteristic dimension of pollution source data, F=65, W1 is the first weight term coefficient, b1 is the first bias term coefficient, ⊙ is element multiplication, M is the mask matrix, and the elements in the mask matrix are 0 or 1.

9. A debris flow monitoring and early warning system using the method described in any one of claims 1 to 8, characterized in that: It includes a monitoring area determination unit, a data acquisition unit, a deformation calculation unit and a monitoring and early warning unit; The monitoring area determination unit is used to construct a debris flow prone area determination model, calculate the debris flow prone area determination value based on rainfall forecast data, debris flow distribution historical data, human production data and the debris flow prone area determination model, and determine the key areas of the debris flow prone area according to the debris flow prone area determination value; The data acquisition unit is used to arrange the sensor network according to the key areas of the debris flow prone area and obtain the multi-index sensor data of the monitoring area; The deformation calculation unit is used to obtain high-resolution satellite remote sensing image data of the monitoring area according to the key areas of the debris flow prone area, and based on the high-resolution satellite remote sensing image data of the monitoring area, the least square method and singular value decomposition method are used to obtain the deformation value of the monitoring area; The monitoring and early warning unit is used to construct a prediction model for debris flow in the monitoring area, obtain the prediction results of debris flow in the monitoring area based on the prediction model for debris flow in the monitoring area, the multi-indicator sensor data of the monitoring area and the deformation value of the monitoring area, and perform early warning operations on the monitoring area based on the prediction results of debris flow in the monitoring area.

Images