A method and system for predicting the probability of debris flow occurrence.

Abstract

This application proposes a method and system for predicting the probability of debris flow occurrence. The method includes: acquiring geological data and rainfall data for various mountainous areas during the prediction period, as well as historical rainfall data for the corresponding periods of the prediction period; determining various geological prediction subsequences for each mountainous area during the prediction period based on the geological data; determining various rainfall prediction subsequences for each mountainous area during the prediction period based on the rainfall data and historical rainfall data for the corresponding periods of the prediction period; and determining the probability of debris flow occurrence for each mountainous area based on the various geological prediction subsequences and the various rainfall prediction subsequences for each mountainous area during the prediction period. The technical solution proposed in this application improves the accuracy of debris flow probability prediction when disaster data is scarce.

Description

Technical Field

[0001] This application relates to the field of debris flow prevention technology, and in particular to a method and system for predicting the probability of debris flow occurrence. Background Technology

[0002] To effectively predict debris flow disasters, existing early warning methods require in-depth analysis of the changing patterns of debris flow characteristics, and are limited by the need for a large amount of debris flow disaster data. However, when equipment acquires debris flow data, it mostly consists of normal, disaster-free data, making debris flow disaster data relatively scarce. Under conditions of scarce debris flow disaster data, the accuracy of predicting debris flows using existing models is low. Summary of the Invention

[0003] This application provides a method and system for predicting the probability of debris flow occurrence, in order to at least solve the technical problem of low accuracy in predicting debris flows using existing models when debris flow disaster data is scarce.

[0004] The first aspect of this application proposes a method for predicting the probability of debris flow occurrence, the method comprising:

[0005] Obtain geological and rainfall data for various mountainous areas during the predicted period, as well as historical rainfall data corresponding to the predicted period.

[0006] Based on the geological data of each region in the mountainous area during the prediction period, determine the geological prediction subsequences of each region in the mountainous area during the prediction period.

[0007] Based on the rainfall data of each area in the mountainous region during the forecast period, and the rainfall data of the corresponding period in previous years, the rainfall prediction subsequences for each area in the mountainous region during the forecast period are determined.

[0008] The probability of debris flow occurrence in each area of ​​the mountainous region is determined based on the geological prediction subsequences and rainfall prediction subsequences for each area of ​​the mountainous region during the predicted period.

[0009] Preferably, determining the geological prediction subsequences for each region of the mountainous area during the prediction period based on geological data of each region during the prediction period includes:

[0010] The geological data of each region in the mountainous area during the prediction period are divided into equal time periods according to a first preset duration to obtain a set of geological data subsequences for each region during the prediction period.

[0011] The geological data subsequence sets for each region during the period to be predicted are input into the pre-trained geological change prediction model to obtain the geological prediction subsequences for each region during the period to be predicted.

[0012] Furthermore, the training process of the geological change prediction model includes:

[0013] Step F1: Obtain geological data subsequences from various historical time periods, then divide the geological data subsequences from various historical time periods into training sets and test sets, and configure the initial parameter set and iteration number threshold of the neural network model;

[0014] Step F2: Extract the i-th geological data subsequence and the (i+1)-th geological data subsequence from the geological data subsequences in the r-th historical period of the training set. Use the i-th geological data subsequence as the input matrix and the (i+1)-th geological data subsequence as the output matrix to train the neural network model. Then, use the test set to verify the trained neural network model to obtain the i-th optimized neural network model in the r-th iteration. Determine whether i is equal to I-1. If yes, proceed to step F4; otherwise, proceed to step F3.

[0015] Step F3: Based on the geological prediction subsequence output by the optimized neural network model in the r-th iteration, solve the objective function. The objective parameter set is obtained, where Q' is the objective parameter set, x i Let y be the i-th geological data subsequence in the geological data subsequence set. i Let I be the geological prediction subsequence output by the neural network model after the (i-1)th geological data subsequence is processed, where I is the number of geological data subsequences in a historical period, and ||·||2 is the L2 norm.

[0016] Step F4: Determine if r equals R. If yes, proceed to step F5; otherwise, set r = r + 1, update the initial parameter set to the target parameter set, and return to step F2.

[0017] Step F5: When the number of iterations equals the iteration threshold, the iteration stops, and the trained geological change prediction model is obtained.

[0018] Furthermore, the step of determining the predicted rainfall subsequences for each region in the mountainous area during the predicted period based on rainfall data for each region in the mountainous area during the predicted period and rainfall data for the corresponding periods in previous years includes:

[0019] The total rainfall for the period to be predicted is determined based on the rainfall data of the corresponding periods in previous years.

[0020] The total rainfall for the period to be predicted is divided into equal time periods according to a first preset duration, so as to obtain the total rainfall for each region in each time period within the period to be predicted.

[0021] The rainfall prediction subsequences for the period to be predicted are determined based on the total rainfall for the period to be predicted and the total rainfall for each segmented period within the period to be predicted.

[0022] Furthermore, determining the probability of debris flow occurrence in each area of ​​the mountainous region based on the geological prediction subsequences of each area during the forecast period and the rainfall prediction subsequences of each area during the forecast period includes:

[0023] By associating the geological prediction subsequences of each region with their corresponding regions and corresponding rainfall prediction subsequences during the period to be predicted, the geological prediction subsequences and rainfall prediction subsequences of each region under each time period segmentation during the period to be predicted are obtained.

[0024] The probability of debris flow occurrence in each mountainous area is determined based on the geological prediction subsequence and rainfall prediction subsequence of each region within the predicted time period.

[0025] Furthermore, the formula for calculating the probability of debris flow occurrence in each area of ​​the mountainous region is as follows:

[0026]

[0027] In the formula, Q i Let P be the probability of a debris flow occurring in the i-th region of the mountainous area. i,t Let be the probability of a debris flow occurring in the i-th region of the mountainous area during the t-th segmentation period, and T be the total number of segmentation periods into which the predicted period is divided according to a first preset duration. Let be the predicted rainfall subsequence for the t-th segment. Let ω be the geological prediction subsequence for the t-th segment. 降雨 ω is the weighting coefficient of rainfall on the probability of debris flow occurrence. 地质 is the weighting coefficient of geological data on the probability of debris flow occurrence, and b is the bias value.

[0028] Furthermore, the method also includes:

[0029] The difference between the geological data subsequence and its corresponding geological prediction subsequence in the geological data subsequence set for each region during the period to be predicted is determined respectively;

[0030] Based on the aforementioned differences, the sum of the differences for each region during the period to be predicted is determined, and the sum of the differences for each region during the period to be predicted is taken as the geological change value for each region.

[0031] When the geological change value is less than or equal to the preset lower limit threshold for geological change, the area corresponding to the geological change value is determined to be a zone of slight geological change.

[0032] When the geological change value is greater than the preset lower threshold of geological change and less than or equal to the preset upper threshold of geological change, the area corresponding to the geological change value is determined to be a moderate geological change area.

[0033] When the geological change value is greater than the preset upper limit threshold for geological change, the area corresponding to the geological change value is determined to be a severely geologically changed area.

[0034] Furthermore, the method also includes:

[0035] An alarm is triggered when the probability of a debris flow occurs exceeds a preset probability threshold.

[0036] A second aspect of this application provides a system for predicting the probability of debris flow occurrence, comprising:

[0037] The acquisition module is used to acquire geological data and rainfall data of various areas in the mountainous region during the period to be predicted, as well as rainfall data of the corresponding period in previous years.

[0038] The first determining module is used to determine the geological prediction sub-sequences of each area in the mountainous region during the prediction period based on the geological data of each area in the mountainous region during the prediction period.

[0039] The second determining module is used to determine the rainfall prediction subsequences for each area of ​​the mountainous region during the prediction period based on the rainfall data of each area in the mountainous region during the prediction period and the rainfall data of the corresponding period in previous years.

[0040] The third determining module is used to determine the probability of debris flow occurrence in each area of ​​the mountainous region based on the geological prediction subsequences of each area in the predicted period and the rainfall prediction subsequences of each area in the mountainous region in the predicted period.

[0041] Preferably, the first determining module is further configured to:

[0042] The geological data of each region in the mountainous area during the prediction period are divided into equal time periods according to a first preset duration to obtain a set of geological data subsequences for each region during the prediction period.

[0043] The geological data subsequence sets for each region during the period to be predicted are input into the pre-trained geological change prediction model to obtain the geological prediction subsequences for each region during the period to be predicted.

[0044] The technical solutions provided by the embodiments of this application have at least the following beneficial effects:

[0045] This application proposes a method and system for predicting the probability of debris flow occurrence. The method includes: acquiring geological data and rainfall data for various mountainous areas during the prediction period, as well as historical rainfall data corresponding to the prediction period; determining various geological prediction subsequences for each mountainous area during the prediction period based on the geological data; determining various rainfall prediction subsequences for each mountainous area during the prediction period based on the rainfall data and historical rainfall data corresponding to the prediction period; and determining the probability of debris flow occurrence for each mountainous area based on the various geological prediction subsequences and the various rainfall prediction subsequences. The technical solution proposed in this application improves the accuracy of debris flow probability prediction in situations where disaster data is scarce.

[0046] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0047] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0048] Figure 1 A flowchart illustrating a method for predicting the probability of debris flow occurrence according to an embodiment of this application;

[0049] Figure 2 This is a first structural diagram of a debris flow occurrence probability prediction system provided according to an embodiment of this application;

[0050] Figure 3 This is a second structural diagram of a debris flow occurrence probability prediction system provided according to an embodiment of this application;

[0051] Figure 4 This is a third structural diagram of a debris flow occurrence probability prediction system provided according to an embodiment of this application. Detailed Implementation

[0052] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0053] This application proposes a method and system for predicting the probability of debris flow occurrence. The method includes: acquiring geological data and rainfall data for various mountainous areas during the prediction period, as well as historical rainfall data corresponding to the prediction period; determining various geological prediction subsequences for each mountainous area during the prediction period based on the geological data; determining various rainfall prediction subsequences for each mountainous area during the prediction period based on the rainfall data and historical rainfall data corresponding to the prediction period; and determining the probability of debris flow occurrence for each mountainous area based on the various geological prediction subsequences and the various rainfall prediction subsequences. The technical solution proposed in this application improves the accuracy of debris flow probability prediction in situations where disaster data is scarce.

[0054] The following describes a method and system for predicting the probability of debris flow occurrence according to an embodiment of this application, with reference to the accompanying drawings.

[0055] Example 1

[0056] Figure 1 This is a flowchart illustrating a method for predicting the probability of debris flow occurrence according to an embodiment of this application, as shown below. Figure 1 As shown, the method includes:

[0057] Step 1: Obtain geological data and rainfall data for each area in the mountainous region during the predicted period, as well as historical rainfall data corresponding to the predicted period.

[0058] It should be noted that geological data of various areas in the mountainous region were obtained; the geological data was collected once at intervals T during disaster-free periods; the geological data were arranged in chronological order to obtain a geological data sequence.

[0059] Define the period to be predicted, the start time of the period to be predicted, and the reference period. The month before, the current month, and the next month before the start time are taken as the period to be predicted. The months corresponding to the period to be predicted in previous years are taken as the reference period.

[0060] Rainfall data for a reference period can be extracted from the above rainfall data. Using the rainfall data for the reference period, the total annual rainfall for the reference period can be obtained.

[0061] Step 2: Determine the geological prediction subsequences for each region in the mountainous area during the prediction period based on the geological data of each region in the mountainous area during the prediction period;

[0062] In this embodiment of the disclosure, step 2 specifically includes:

[0063] The geological data of each region in the mountainous area during the prediction period are divided into equal time periods according to a first preset duration to obtain a set of geological data subsequences for each region during the prediction period.

[0064] The geological data subsequence sets for each region during the period to be predicted are input into the pre-trained geological change prediction model to obtain the geological prediction subsequences for each region during the period to be predicted.

[0065] It should be noted that the training process of the geological change prediction model includes:

[0066] Step F1: Obtain geological data subsequences from various historical time periods, then divide the geological data subsequences from various historical time periods into training sets and test sets, and configure the initial parameter set and iteration number threshold of the neural network model;

[0067] Step F2: Extract the i-th geological data subsequence and the (i+1)-th geological data subsequence from the geological data subsequences in the r-th historical period of the training set. Use the i-th geological data subsequence as the input matrix and the (i+1)-th geological data subsequence as the output matrix to train the neural network model. Then, use the test set to verify the trained neural network model to obtain the i-th optimized neural network model in the r-th iteration. Determine whether i is equal to I-1. If yes, proceed to step F4; otherwise, proceed to step F3.

[0068] Step F3: Based on the geological prediction subsequence output by the optimized neural network model in the r-th iteration, solve the objective function. The objective parameter set is obtained, where Q' is the objective parameter set, x i Let y be the i-th geological data subsequence in the geological data subsequence set. i Let I be the geological prediction subsequence output by the neural network model after the (i-1)th geological data subsequence is processed, where I is the number of geological data subsequences in a historical period, and ||·||2 is the L2 norm.

[0069] Step F4: Determine if r equals R. If yes, proceed to step F5; otherwise, set r = r + 1, update the initial parameter set to the target parameter set, and return to step F2.

[0070] Step F5: When the number of iterations equals the iteration threshold, the iteration stops, and the trained geological change prediction model is obtained.

[0071] It should be noted that the parameters of the neural network model are initialized, including weights and bias terms, and the iteration number threshold is set, which is the maximum number of iterations for model training, to control the termination condition of the training process; each iteration selects two consecutive subsequences of geological data, the first as input data and the second as the expected output (or label). Input data is forward-propagated through a neural network to obtain prediction results, simulating the actual evolution process of time series data, enabling the model to learn the patterns of geological conditions changing over time. Each time a subset of geological data is traversed, the iteration count is increased, ensuring the model undergoes multiple training cycles to gradually adjust parameters until the preset number of iterations is met. The objective function is calculated using the difference between the output matrix (predicted result) and the expected matrix (actual geological data). The model parameters are adjusted based on the gradient of the objective function to make the prediction results closer to the actual geological data. The model's prediction performance is improved by continuously optimizing the objective function (typically using methods such as gradient descent or other optimization algorithms to update parameters). The current iteration count is checked to see if a preset threshold has been reached. If it has, the model is considered sufficiently trained, and iteration stops; otherwise, training continues until the predetermined number of iterations is reached.

[0072] In this embodiment of the disclosure, the method further includes:

[0073] The difference between the geological data subsequence and its corresponding geological prediction subsequence in the geological data subsequence set for each region during the period to be predicted is determined respectively;

[0074] Based on the aforementioned differences, the sum of the differences for each region during the period to be predicted is determined, and the sum of the differences for each region during the period to be predicted is taken as the geological change value for each region.

[0075] When the geological change value is less than or equal to the preset lower limit threshold for geological change, the area corresponding to the geological change value is determined to be a zone of slight geological change.

[0076] When the geological change value is greater than the preset lower threshold of geological change and less than or equal to the preset upper threshold of geological change, the area corresponding to the geological change value is determined to be a moderate geological change area.

[0077] When the geological change value is greater than the preset upper limit threshold for geological change, the area corresponding to the geological change value is determined to be a severely geologically changed area.

[0078] It should be noted that the geological changes in areas with slight geological changes are not significant, and problems such as loose soil are not likely to occur.

[0079] The geological changes in areas with severe geological changes are significant, and problems such as loose soil and weathered rocks are likely to occur, making them prone to mudslides during rainfall.

[0080] Considering that different levels of geological change have varying impacts on debris flow formation, and thus different probabilities of debris flows occurring under rainfall conditions, this study quantifies the degree of geological change into different levels by setting clear thresholds, facilitating subsequent risk assessment and early warning. The geological change value is obtained by calculating the difference between the actual geological data subsequence and its corresponding geological prediction subsequence; this value reflects the difference between actual and predicted geological data and is used to measure the degree of impact of geological change on debris flow formation. The geological change level is then classified by comparing the geological change value with the preset geological change threshold.

[0081] Step 3: Based on the rainfall data of each area in the mountainous region during the forecast period, and the rainfall data of the corresponding period in previous years, determine the rainfall prediction subsequences for each area in the mountainous region during the forecast period.

[0082] In this embodiment of the disclosure, step 3 specifically includes:

[0083] The total rainfall for the period to be predicted is determined based on the rainfall data of the corresponding periods in previous years.

[0084] The total rainfall for the period to be predicted is divided into equal time periods according to a first preset duration, so as to obtain the total rainfall for each region in each time period within the period to be predicted.

[0085] The rainfall prediction subsequences for the period to be predicted are determined based on the total rainfall for the period to be predicted and the total rainfall for each segmented period within the period to be predicted.

[0086] It should be noted that the period to be predicted, the start time of the period to be predicted, and the reference period are defined. The month before, the current month, and the next month before the start time are used as the period to be predicted; the months corresponding to the period to be predicted in previous years are used as the reference period.

[0087] Rainfall data for the reference period is extracted from the above rainfall data, and the total annual rainfall for the reference period is obtained using the rainfall data for the above reference period.

[0088] Based on the total rainfall for the reference periods mentioned above in each year, the total rainfall for the period to be predicted is determined using the following formula:

[0089]

[0090] In the formula, This indicates the total rainfall for the period to be predicted; This represents the total rainfall during the reference period in year j; n represents the number of reference periods obtained.

[0091] The rainfall amount before the start time within the forecast period is extracted from the above rainfall data. Using the total rainfall for the forecast period and the rainfall amount before the start time, the rainfall amount from the start time to the end time within the forecast period is determined, as shown in the following formula:

[0092]

[0093] In the formula, This represents the rainfall from the start time to the end time within the forecast period; This represents the rainfall amount before the start time within the forecast period;

[0094] The following formula is used to determine the rainfall prediction subsequence from the start time to the end time within the forecast period;

[0095]

[0096] In the formula, The predicted rainfall subsequence from the start time to the end time within the predicted period is the predicted rainfall subsequence of the t-th segmented period; T represents the total number of segmented periods in the predicted period according to the first preset duration; t represents the time from the start time to the prediction time within the predicted period; K0 represents the soil permeability coefficient of the reference period; g(B) represents the reduction factor; a represents a constant; B represents the geological change value.

[0097] The above methods take into account the impact of geological variations on soil permeability, thereby improving the accuracy of rainfall prediction.

[0098] Step 4: Determine the probability of debris flow occurrence in each area of ​​the mountainous region based on the geological prediction subsequences and rainfall prediction subsequences for each area of ​​the mountainous region during the forecast period.

[0099] In this embodiment of the disclosure, step 4 specifically includes:

[0100] By associating the geological prediction subsequences of each region with their corresponding regions and corresponding rainfall prediction subsequences during the period to be predicted, the geological prediction subsequences and rainfall prediction subsequences of each region under each time period segmentation during the period to be predicted are obtained.

[0101] The probability of debris flow occurrence in each mountainous area is determined based on the geological prediction subsequence and rainfall prediction subsequence of each region within the predicted time period.

[0102] The formula for calculating the probability of debris flow occurrence in each area of ​​the mountainous region is as follows:

[0103]

[0104] In the formula, Q i Let P be the probability of a debris flow occurring in the i-th region of the mountainous area. i,t Let be the probability of a debris flow occurring in the i-th region of the mountainous area during the t-th segmentation period, and T be the total number of segmentation periods into which the predicted period is divided according to a first preset duration. Let be the predicted rainfall subsequence for the t-th segment. Let ω be the geological prediction subsequence for the t-th segment. 降雨 ω is the weighting coefficient of rainfall on the probability of debris flow occurrence. 地质 is the weighting coefficient of geological data on the probability of debris flow occurrence, and b is the bias value.

[0105] In this embodiment of the disclosure, the method further includes:

[0106] An alarm is triggered when the probability of a debris flow occurs exceeds a preset probability threshold.

[0107] It should be noted that this embodiment collects geological data from different time periods, reflecting changes in geological conditions over time. Since the geological data is collected periodically during disaster-free periods, it provides a benchmark for comparing geological changes. By dividing the geological data sequence into equal time-period segments, multiple geological data subsets are formed, which helps to better capture the trend of geological conditions changing over time during subsequent training. By setting an objective function, the neural network model is trained using the geological data subsets to predict changes in geological conditions. The objective function aims to minimize the difference between the predicted values ​​and the actual geological data, thereby optimizing the model's predictive ability. Based on the trained geological change prediction model, geological prediction subsets are output, and the geological change level is determined accordingly. Since the geological change level not only affects subsequent rainfall to a certain extent but also soil stability, combining the geological change level and rainfall data can more accurately predict the rainfall change trend in a specific area. Combining the geological prediction subsets with the rainfall prediction subsets can further refine the probability of debris flow occurrence. A probability threshold is set as an early warning standard; when the predicted probability of debris flow occurrence reaches or exceeds this threshold, the system will automatically issue an early warning signal. This mechanism can help take timely preventative measures and reduce losses caused by disasters.

[0108] In summary, the debris flow probability prediction method proposed in this embodiment does not require disaster data at the time of debris flow to predict whether a debris flow will occur within a certain period of time. Instead, it uses normal data from disaster-free periods to predict geological change trends and combines rainfall prediction to assess the probability of debris flow occurrence. This method can improve the accuracy of prediction even when disaster data is scarce. It can not only help identify potential debris flow-prone areas but also provide early warnings.

[0109] Example 2

[0110] Figure 2 This is a structural diagram of a debris flow occurrence probability prediction system according to an embodiment of this application, as shown below. Figure 2 As shown, the system includes:

[0111] The acquisition module 100 is used to acquire geological data and rainfall data of various areas in the mountainous region during the period to be predicted, as well as rainfall data of the corresponding period in previous years.

[0112] The first determining module 200 is used to determine the geological prediction sub-sequences of each area in the mountainous region during the prediction period based on the geological data of each area in the mountainous region during the prediction period.

[0113] The second determining module 300 is used to determine the rainfall prediction subsequences for each region in the mountainous area during the prediction period based on the rainfall data of each region in the mountainous area during the prediction period and the rainfall data of the corresponding period in previous years.

[0114] The third determining module 400 is used to determine the probability of debris flow occurrence in each area of ​​the mountainous region based on the geological prediction subsequences of each area in the predicted period and the rainfall prediction subsequences of each area in the mountainous region in the predicted period.

[0115] In this embodiment of the disclosure, the first determining module 200 is further configured to:

[0116] The geological data of each region in the mountainous area during the prediction period are divided into equal time periods according to a first preset duration to obtain a set of geological data subsequences for each region during the prediction period.

[0117] The geological data subsequence sets for each region during the period to be predicted are input into the pre-trained geological change prediction model to obtain the geological prediction subsequences for each region during the period to be predicted.

[0118] It should be noted that the first determining module 200 is also used to train the geological change prediction model;

[0119] The training process of the geological change prediction model includes:

[0120] Step E1: Obtain geological data subsequences from various historical time periods, then divide the geological data subsequences from various historical time periods into training sets and test sets, and configure the initial parameter set and iteration number threshold of the neural network model;

[0121] Step E2: Extract the i-th geological data subsequence and the (i+1)-th geological data subsequence from the geological data subsequences in the r-th historical period of the training set. Use the i-th geological data subsequence as the input matrix and the (i+1)-th geological data subsequence as the output matrix to train the neural network model. Then, use the test set to verify the trained neural network model to obtain the i-th optimized neural network model in the r-th iteration. Determine whether i is equal to I-1. If yes, proceed to step E4; otherwise, proceed to step E3.

[0122] Step E3: Based on the geological prediction subsequence output by the optimized neural network model in the r-th iteration, solve the objective function. The objective parameter set is obtained, where Q' is the objective parameter set, x i Let y be the i-th geological data subsequence in the geological data subsequence set. i Let I be the geological prediction subsequence output by the neural network model after the (i-1)th geological data subsequence is processed, where I is the number of geological data subsequences in a historical period, and ||·||2 is the L2 norm.

[0123] Step E4: Determine if r equals R. If yes, proceed to step F5; otherwise, set r = r + 1, update the initial parameter set to the target parameter set, and return to step E2.

[0124] Step E5: When the number of iterations equals the iteration threshold, the iteration stops, and the trained geological change prediction model is obtained.

[0125] In this embodiment of the disclosure, the second determining module 300 is further configured to:

[0126] The total rainfall for the period to be predicted is determined based on the rainfall data of the corresponding periods in previous years.

[0127] The total rainfall for the period to be predicted is divided into equal time periods according to a first preset duration, so as to obtain the total rainfall for each region in each time period within the period to be predicted.

[0128] The rainfall prediction subsequences for the period to be predicted are determined based on the total rainfall for the period to be predicted and the total rainfall for each segmented period within the period to be predicted.

[0129] In this embodiment of the disclosure, the third determining module 400 is further configured to:

[0130] By associating the geological prediction subsequences of each region with their corresponding regions and corresponding rainfall prediction subsequences during the period to be predicted, the geological prediction subsequences and rainfall prediction subsequences of each region under each time period segmentation during the period to be predicted are obtained.

[0131] The probability of debris flow occurrence in each mountainous area is determined based on the geological prediction subsequence and rainfall prediction subsequence of each region within the predicted time period.

[0132] The formula for calculating the probability of debris flow occurrence in each area of ​​the mountainous region is as follows:

[0133]

[0134] In the formula, Q i Let P be the probability of a debris flow occurring in the i-th region of the mountainous area. i,t Let be the probability of a debris flow occurring in the i-th region of the mountainous area during the t-th segmentation period, and T be the total number of segmentation periods into which the predicted period is divided according to a first preset duration. Let be the predicted rainfall subsequence for the t-th segment. Let ω be the geological prediction subsequence for the t-th segment. 降雨 ω is the weighting coefficient of rainfall on the probability of debris flow occurrence. 地质 is the weighting coefficient of geological data on the probability of debris flow occurrence, and b is the bias value.

[0135] In the embodiments disclosed herein, such as Figure 3 As shown, the system further includes: a judgment module 500;

[0136] The judgment module 500 is used for:

[0137] The difference between the geological data subsequence and its corresponding geological prediction subsequence in the geological data subsequence set for each region during the period to be predicted is determined respectively;

[0138] Based on the aforementioned differences, the sum of the differences for each region during the period to be predicted is determined, and the sum of the differences for each region during the period to be predicted is taken as the geological change value for each region.

[0139] When the geological change value is less than or equal to the preset lower limit threshold for geological change, the area corresponding to the geological change value is determined to be a zone of slight geological change.

[0140] When the geological change value is greater than the preset lower threshold of geological change and less than or equal to the preset upper threshold of geological change, the area corresponding to the geological change value is determined to be a moderate geological change area.

[0141] When the geological change value is greater than the preset upper limit threshold for geological change, the area corresponding to the geological change value is determined to be a severely geologically changed area.

[0142] In the embodiments disclosed herein, such as Figure 4 As shown, the system also includes: an alarm module 600;

[0143] The alarm module 600 is used for:

[0144] An alarm is triggered when the probability of a debris flow occurs exceeds a preset probability threshold.

[0145] In summary, the debris flow occurrence probability prediction system proposed in this embodiment improves the accuracy of debris flow occurrence probability prediction when disaster data is scarce.

[0146] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0147] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0148] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for predicting the probability of debris flow occurrence, characterized in that, The method includes: Obtain geological and rainfall data for various mountainous areas during the predicted period, as well as historical rainfall data corresponding to the predicted period. Based on the geological data of each region in the mountainous area during the prediction period, determine the geological prediction subsequences of each region in the mountainous area during the prediction period. Based on the rainfall data of each area in the mountainous region during the forecast period, and the rainfall data of the corresponding period in previous years, the rainfall prediction subsequences for each area in the mountainous region during the forecast period are determined. The probability of debris flow occurrence in each area of ​​the mountainous region is determined based on the geological prediction subsequences and rainfall prediction subsequences for each area of ​​the mountainous region during the forecast period. The step of determining the geological prediction sub-sequences for each region of the mountainous area during the predicted period based on geological data of each region during the predicted period includes: The geological data of each region in the mountainous area during the prediction period are divided into equal time periods according to a first preset duration to obtain a set of geological data subsequences for each region during the prediction period. The geological data subsequence sets of each region during the period to be predicted are input into the pre-trained geological change prediction model to obtain the geological prediction subsequences of each region during the period to be predicted. The training process of the geological change prediction model includes: Step F1: Obtain geological data subsequences from various historical time periods, then divide the geological data subsequences from various historical time periods into training sets and test sets, and configure the initial parameter set and iteration number threshold of the neural network model; Step F2: Extract the i-th and (i+1)-th geological data subsequences from the geological data subsequences of each historical period in the training set. Use the i-th geological data subsequence as the input matrix and the (i+1)-th geological data subsequence as the output matrix to train the neural network model. Then, use the test set to validate the trained neural network model, obtaining the i-th optimized neural network model in the r-th iteration. Determine whether i equals... If yes, proceed to step F4; otherwise, proceed to step F3. Step F3: Based on the geological prediction subsequence output by the optimized neural network model in the r-th iteration, solve the objective function. The target parameter set is obtained, where, For the target parameter set, For the first subsequence of geological data A geological data subsequence, For the first The geological prediction subsequence output by the neural network model after processing a subsequence of geological data. The number of geological data subsequences within a historical period. It is a norm 2; Step F4: Determine if r equals R. If yes, proceed to step F5; otherwise, set r = r + 1, update the initial parameter set to the target parameter set, and return to step F2. Step F5: When the number of iterations equals the iteration threshold, the iteration stops, and the trained geological change prediction model is obtained; The step of determining the rainfall prediction sub-sequences for each region in the mountainous area during the predicted period based on rainfall data for each region in the mountainous area during the predicted period and rainfall data for the corresponding periods in previous years includes: The total rainfall for the period to be predicted is determined based on the rainfall data of the corresponding periods in previous years. The total rainfall for the period to be predicted is divided into equal time periods according to a first preset duration, so as to obtain the total rainfall for each region in each time period within the period to be predicted. The rainfall prediction subsequence within the predicted period is determined based on the total rainfall during the predicted period and the total rainfall during each segmented period within the predicted period. The step of determining the probability of debris flow occurrence in each area of ​​the mountainous region based on the geological prediction subsequences and rainfall prediction subsequences for each area of ​​the mountainous region during the forecast period includes: By associating the geological prediction subsequences of each region with their corresponding regions and corresponding rainfall prediction subsequences during the period to be predicted, the geological prediction subsequences and rainfall prediction subsequences of each region under each time period segmentation during the period to be predicted are obtained. The probability of debris flow occurrence in each mountainous area is determined based on the geological prediction subsequence and rainfall prediction subsequence of each area under the time segmentation of each equal time period within the predicted time period. The formulas for calculating the probability of debris flow occurrence in various areas of the mountainous region are as follows: In the formula, Let be the probability of a debris flow occurring in the i-th region of the mountainous area. Let be the probability of a debris flow occurring in the i-th region of the mountainous area during the t-th time segment. The total number of time periods to be predicted, which are divided into equal time periods according to a first preset duration, wherein, , Let be the predicted rainfall subsequence for the t-th segment. Let be the geological prediction subsequence for the t-th segment. This represents the weighting coefficient of rainfall on the probability of debris flow occurrence. This represents the weighting coefficient of geological data on the probability of debris flow occurrence. This is the bias value.

2. The method as described in claim 1, characterized in that, The method further includes: The difference between the geological data subsequence and its corresponding geological prediction subsequence in the geological data subsequence set for each region during the period to be predicted is determined respectively; Based on the aforementioned differences, the sum of the differences for each region during the period to be predicted is determined, and the sum of the differences for each region during the period to be predicted is taken as the geological change value for each region. When the geological change value is less than or equal to the preset lower limit threshold for geological change, the area corresponding to the geological change value is determined to be a zone of slight geological change. When the geological change value is greater than the preset lower threshold of geological change and less than or equal to the preset upper threshold of geological change, the area corresponding to the geological change value is determined to be a moderate geological change area. When the geological change value is greater than the preset upper limit threshold for geological change, the area corresponding to the geological change value is determined to be a severely geologically changed area.

3. The method as described in claim 2, characterized in that, The method further includes: An alarm is triggered when the probability of a mudslide occurs exceeds a preset probability threshold.

4. A system for predicting the probability of debris flow occurrence, characterized in that, The system implements the method as described in claim 1, the system comprising: The acquisition module is used to acquire geological data and rainfall data of various areas in the mountainous region during the period to be predicted, as well as rainfall data of the corresponding period in previous years. The first determining module is used to determine the geological prediction sub-sequences of each area in the mountainous region during the prediction period based on the geological data of each area in the mountainous region during the prediction period. The second determining module is used to determine the rainfall prediction subsequences for each area of ​​the mountainous region during the prediction period based on the rainfall data of each area in the mountainous region during the prediction period and the rainfall data of the corresponding period in previous years. The third determining module is used to determine the probability of debris flow occurrence in each area of ​​the mountainous region based on the geological prediction subsequences of each area in the predicted period and the rainfall prediction subsequences of each area in the mountainous region in the predicted period.

Images