Landslide prediction accuracy improvement method and device, computer equipment and storage medium

Abstract

The embodiment of the application discloses a landslide prediction accuracy improving method and device, computer equipment and a storage medium. The method comprises the following steps: adding a derived feature obtained from experience knowledge into a landslide prediction model to form a new feature; and training the landslide prediction model by using the new feature, so as to perform landslide prediction by using the trained landslide prediction model. The method can improve the prediction accuracy of the landslide prediction model.

Description

TECHNICAL FIELD

[0001] The present application relates to a landslide prediction method, more particularly to a landslide prediction accuracy improvement method and device, a computer device and a storage medium. BACKGROUND

[0002] Landslide is a common natural disaster, which has a great impact on human life, so the identification and classification of landslide is a key problem for disaster prevention and post-disaster assessment. The existing landslide prediction method is based on optical image mountain landslide extraction, which mainly considers spectral, texture, shape and morphological features. The machine learning model constructed from these features is used for landslide prediction. The machine learning model constructs multi-dimensional features for each training sample, including spectral, texture and geometric features, and determines the threshold of each dimension feature through training sample learning.

[0003] However, the existing landslide prediction method only considers spectral, texture and geometric features, and does not consider the relationship between features and scenes. The effectiveness of these features is not very high, which may lead to low accuracy of the constructed model in predicting landslides.

[0004] Therefore, it is necessary to design a new method to improve the prediction accuracy of the landslide prediction model. SUMMARY

[0005] The present application aims to overcome the defects of the prior art and provide a landslide prediction accuracy improvement method, device, computer device and storage medium.

[0006] To achieve the above-mentioned purpose, the present application adopts the following technical solution: a landslide prediction accuracy improvement method, comprising:

[0007] Derivative features obtained from empirical knowledge are added to the landslide prediction model to form new features;

[0008] The landslide prediction model is trained using the new features, and the trained landslide prediction model is used for landslide prediction.

[0009] Further technical solutions are as follows: the derivative features include single-variable derivative features in historical time and multi-variable interactive derivative features.

[0010] Further technical solutions are as follows: the single-variable derivative features in historical time include features of shape before and after rainfall.

[0011] Further technical solutions are as follows: the multi-variable interactive derivative features include rock mass freeze-thaw index.

[0012] Further technical solutions are as follows: the derivative features obtained from empirical knowledge are added to the landslide prediction model to form new features, comprising:

[0013] When the derived feature is the feature of the shape before and after the rainfall, the rainfall data corresponding to each slope unit is sorted to obtain a sorting result;

[0014] The shape before and after the rainfall of the sorting result is determined according to the set index matching type;

[0015] The features of the shape before and after the rainfall within a day, the features of the shape before and after the rainfall within a week and the features of the shape before and after the rainfall within a month are added into the landslide prediction model according to time periods.

[0016] A further technical solution is that the derived feature obtained from the experience knowledge is added into the landslide prediction model to form a new feature, including:

[0017] When the derived feature is the rock mass freeze-thaw index, the rock mass freeze-thaw index is simplified to obtain a processing result;

[0018] The processing result is added into the landslide prediction model to obtain a new feature.

[0019] A further technical solution is that when the derived feature is the rock mass freeze-thaw index, the rock mass freeze-thaw index is simplified to obtain a processing result, including:

[0020] When the derived feature is the rock mass freeze-thaw index, the temperature corresponding to each slope unit within a preset time period is extracted, and the diurnal difference is divided to obtain a temperature difference;

[0021] It is judged whether the temperature difference is greater than a preset threshold value;

[0022] If the temperature difference is greater than the preset threshold value, the freeze-thaw frequency of the rock mass is calculated, and the ratio value of the freeze-thaw frequency of the rock mass and the related constant of the rock mass is taken as a simplified index to obtain a processing result.

[0023] If the temperature difference is not greater than the preset threshold value, the temperature corresponding to each slope unit within a preset time period is extracted, and the diurnal difference is divided to obtain a temperature difference when the derived feature is the rock mass freeze-thaw index.

[0024] The application also provides a landslide prediction accuracy improving device, including:

[0025] An adding unit is configured to add a derived feature obtained from experience knowledge into a landslide prediction model to form a new feature;

[0026] A training unit is configured to train the landslide prediction model by using the new feature, so that the landslide prediction model is trained to predict landslides.

[0027] The application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to realize the method.

[0028] The application further provides a storage medium, which stores a computer program, and the computer program is executed by a processor to realize the method.

[0029] Compared with the prior art, the application has the beneficial effects that: the application adds the derived features obtained from the experience knowledge into the landslide prediction model to improve the feature effectiveness of the landslide prediction model, and then trains the landslide prediction model by using the new features to realize the landslide prediction by using the trained landslide prediction model, thereby improving the prediction accuracy of the landslide prediction model.

[0030] The application will be further described below in combination with the drawings and specific embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0031] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings needed in the embodiment description will be briefly introduced below. Obviously, the drawings in the embodiment description are some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0032] Figure 1 The flowchart of the landslide prediction accuracy improving method provided by the embodiment of the application is shown in the figure.

[0033] Figure 2 The sub-flowchart of the landslide prediction accuracy improving method provided by the embodiment of the application is shown in the figure.

[0034] Figure 3 The sub-flowchart of the landslide prediction accuracy improving method provided by the embodiment of the application is shown in the figure.

[0035] Figure 4 The sub-flowchart of the landslide prediction accuracy improving method provided by the embodiment of the application is shown in the figure.

[0036] Figure 5 The schematic block diagram of the landslide prediction accuracy improving device provided by the embodiment of the application is shown in the figure.

[0037] Figure 6 The schematic block diagram of the computer device provided by the embodiment of the application is shown in the figure. DETAILED DESCRIPTION

[0038] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be clearly and completely described in the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of the present application.

[0039] It should be understood that the terms "comprising" and "including" as used in the specification and the appended claims indicate the presence of the described features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0040] It should also be understood that the terms used in the present application specification are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the present application specification and the appended claims, the singular forms "a", "an" and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0041] It should be further understood that the term "and / or" used in the present application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations thereof, and includes these combinations.

[0042] Please refer to Figure 1 , Figure 1 The schematic flowchart of the landslide prediction accuracy improving method provided by the embodiments of the present application is shown in FIG. 1. The landslide prediction accuracy improving method is applied in a server. By fully considering the characteristics of features and scenes in the modeling process and adding rich experience knowledge, if effective information can be extracted from the original data, the accuracy of model prediction can be greatly improved.

[0043] Figure 1 The flowchart of the landslide prediction accuracy improving method provided by the embodiments of the present application is shown in FIG. 1. As shown in FIG. 1, the method comprises the following steps S110 to S120. Figure 1

[0044] S110, adding derived features obtained from experience knowledge in the landslide prediction model to form new features.

[0045] In the present embodiment, the new features refer to the features formed after adding the derived features obtained from experience knowledge in the landslide prediction model.

[0046] In the present embodiment, the derived features include single-variable derived features in historical time and multi-variable interaction derived features.

[0047] ​Specifically, the derived feature of the single variable over historical time includes a feature of rainfall pattern before and after rainfall. According to empirical knowledge, rainfall conditions can be divided into four types: large before and small after, with a landslide frequency of 16.6%; large before and large after, with a landslide frequency of 17%; large in the middle, with a landslide frequency of 11%; and small before and large after, with a landslide frequency of 55.4%. Therefore, the rainfall pattern before and after rainfall can be extracted in a period of time, such as one day, one week, or one month, and is marked as 1-4 types.

[0048] In addition, the derived feature of the multi-variable interaction includes a rock mass freeze-thaw index. The relative association between different features can mine the specific meaning of the object in the scene. For example, air temperature generally has little effect on landslides, but in recent years, the sudden rise and fall of ground heat and winter temperature can easily cause freeze-thaw phenomena in rock mass, and the destruction of rock mass will increase the probability of landslides.

[0049] In an embodiment, referring to Figure 2 The step S110 described above can include steps S111-S113.

[0050] S111, when the derived feature is the feature of the rainfall pattern before and after rainfall, the rainfall data corresponding to each slope unit is sorted to obtain a sorting result.

[0051] In the embodiment, the sorting result refers to a result obtained by sorting the rainfall data corresponding to each slope unit in a set rule, such as from large to small.

[0052] S112, determining the rainfall pattern before and after rainfall of the sorting result according to a set index matching type;

[0053] S113, adding the features of the rainfall pattern before and after rainfall in one day, the features of the rainfall pattern before and after rainfall in one week, and the features of the rainfall pattern before and after rainfall in one month into the landslide prediction model according to time periods.

[0054] Specifically, the rainfall in one month is extracted for feature extraction, the rainfall data corresponding to each slope unit is sorted, and the corresponding index matching type is determined, such as decreasing for large before and small after, and maximum value not at the beginning and end for large in the middle. According to time periods, three new features can be added to the model, including the rainfall pattern before and after rainfall in one day, the rainfall pattern before and after rainfall in one week, and the rainfall pattern before and after rainfall in one month, which can increase the coverage of long-term rainfall information in the model.

[0055] In the embodiment, the index matching type refers to the rainfall type corresponding to the sorting result of the rainfall data.

[0056] In an embodiment, referring to Figure 3 The step S110 described above can include steps S110a-S110b.

[0057] S110a, when the derived feature is the rock mass freeze-thaw index, performing a simplified processing on the rock mass freeze-thaw index to obtain a processing result.

[0058] In this embodiment, the processing result includes the index content obtained after the simplified processing on the rock mass freeze-thaw index.

[0059] In an embodiment, referring to Figure 4 The step S110a described above can include S110a1-S110a3.

[0060] S110a1, when the derived feature is the rock mass freeze-thaw index, extracting the temperature corresponding to each slope unit in a preset time period and performing a day-night difference to obtain a temperature difference.

[0061] In this embodiment, the temperature difference refers to the day-night difference of the temperature corresponding to each slope unit in a preset time period.

[0062] S110a2, judging whether the temperature difference is greater than a preset threshold value.

[0063] S110a3, if the temperature difference is greater than the preset threshold value, calculating the freeze-thaw frequency of the rock mass, and taking the ratio value of the freeze-thaw frequency of the rock mass and the relevant constant of the rock mass as a simplified index to obtain the processing result.

[0064] In this embodiment, the relevant constant of the rock mass includes axial stress, elastic modulus, load force, etc.

[0065] If the temperature difference is not greater than the preset threshold value, the step S110a1 is performed.

[0066] Since the freeze-thaw index is mostly generated by indoor test and numerical simulation, it is difficult to obtain the crack geometry and spatial position of regional rock mass, rock water content and freezing speed, etc. in practice, so the freeze-thaw index is simplified. Specifically, the temperature corresponding to each slope unit is extracted for a period of time, such as one month, and the day-night difference is obtained. If the temperature difference is greater than a certain threshold value, the freeze-thaw frequency is calculated, and the ratio value of the freeze-thaw frequency and the relevant constant of the rock mass is taken as a simplified index. The rock mass freeze-thaw index can increase the information amount covering the extreme climate in the model.

[0067] S110b, adding the processing result to the landslide prediction model to obtain a new feature.

[0068] S120, training the landslide prediction model using the new feature to perform landslide prediction using the trained landslide prediction model.

[0069] The derived features obtained by experience knowledge are added, the effectiveness of corresponding features of the model is improved, effective feature extraction can significantly improve the model quality, improve evaluation indexes such as mean square error, accuracy, recall rate and the like, and can further enhance the explainability of the model and the trust degree of customers to the model.

[0070] The landslide prediction accuracy improving method improves the feature effectiveness of the landslide prediction model by adding the derived features obtained by experience knowledge into the landslide prediction model, trains the landslide prediction model by using the new features, performs landslide prediction by using the trained landslide prediction model, and realizes improving the prediction accuracy of the landslide prediction model.

[0071] Figure 5 is a schematic block diagram of a landslide prediction accuracy improving device 300 provided by an embodiment of the present application. As shown in Figure 5 corresponding to the above landslide prediction accuracy improving method, the present application further provides a landslide prediction accuracy improving device 300. The landslide prediction accuracy improving device 300 includes units for executing the above landslide prediction accuracy improving method, and the device can be in a server. Specifically, please refer to Figure 5 , the landslide prediction accuracy improving device 300 includes an adding unit 301 and a training unit 302.

[0072] The adding unit 301 is used for adding derived features obtained by experience knowledge into the landslide prediction model to form new features; and the training unit 302 is used for training the landslide prediction model by using the new features, and performing landslide prediction by using the trained landslide prediction model.

[0073] In an embodiment, the adding unit 301 includes a sorting sub-unit, a form determining sub-unit and a feature adding sub-unit.

[0074] The sorting sub-unit is used for sorting the rainfall data corresponding to each slope unit when the derived feature is the feature of the form before and after rainfall, to obtain a sorting result; the form determining sub-unit is used for determining the form before and after rainfall of the sorting result according to a set index matching type; and the feature adding sub-unit is used for adding the features of the form before and after rainfall within one day, the features of the form before and after rainfall within one week and the features of the form before and after rainfall within one month into the landslide prediction model according to time periods.

[0075] In an embodiment, the adding unit 301 includes a simplification processing sub-unit and a result adding sub-unit.

[0076] The simplification processing sub-unit is used for performing simplification processing on the rock mass freeze-thaw index when the derived feature is the rock mass freeze-thaw index, to obtain a processing result; and the result adding sub-unit is used for adding the processing result into the landslide prediction model to obtain new features.

[0077] In an embodiment, the adding unit 301 comprises a temperature difference calculation subunit, a temperature difference judgment subunit, and an index calculation subunit.

[0078] The temperature difference calculation subunit is configured to, when the derived feature is the rock mass freeze-thaw index, extract the temperature corresponding to each slope unit in a preset time period and perform day-night difference to obtain a temperature difference. The temperature difference judgment subunit is configured to judge whether the temperature difference is greater than a preset threshold. The index calculation subunit is configured to, when the temperature difference is greater than the preset threshold, calculate the number of rock mass freeze-thaw times, and take the ratio value of the number of rock mass freeze-thaw times and the relevant constant of the rock mass as a simple index to obtain a processing result.

[0079] It should be noted that the specific implementation process of the above-mentioned landslide prediction accuracy improving device 300 and each unit can be clearly understood by those skilled in the art, which can be referred to the corresponding description in the foregoing method embodiments, and will not be described here for the convenience and brevity of description.

[0080] The above-mentioned landslide prediction accuracy improving device 300 can be realized in the form of a computer program, which can run on a computer device as shown in the figure. Figure 6 The computer device can be a server.

[0081] Please refer to Figure 6 , Figure 6 is a schematic block diagram of a computer device provided by an embodiment of the present application. The computer device 500 is a server, wherein the server can be a stand-alone server or a server cluster composed of multiple servers.

[0082] Referring to Figure 6 , the computer device 500 comprises a processor 502, a memory, and a network interface 505 connected through a system bus 501, wherein the memory can comprise a non-volatile storage medium 503 and an internal memory 504.

[0083] The non-volatile storage medium 503 can store an operating system 5031 and a computer program 5032. The computer program 5032 comprises program instructions, which, when executed, can cause the processor 502 to perform a landslide prediction accuracy improving method.

[0084] The processor 502 is configured to provide computing and control capabilities to support the operation of the entire computer device 500.

[0085] The memory 504 provides an environment for running the computer program 5032 in the nonvolatile storage medium 503, and the computer program 5032, when executed by the processor 502, can enable the processor 502 to perform a landslide prediction accuracy improvement method.

[0086] The network interface 505 is used for network communication with other devices. Those skilled in the art can understand that, Figure 6 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device 500 to which the scheme of the present application is applied. The specific computer device 500 can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0087] The processor 502 is configured to run the computer program 5032 stored in the memory to implement the following steps:

[0088] The derived features obtained from experience knowledge are added to the landslide prediction model to form new features, and the landslide prediction model is trained using the new features to perform landslide prediction using the trained landslide prediction model.

[0089] The derived features include single-variable derived features over historical time and multi-variable interaction derived features.

[0090] The single-variable derived features over historical time include features of the shape before and after rainfall.

[0091] The multi-variable interaction derived features include rock mass freeze-thaw index.

[0092] In an embodiment, when implementing the step of adding derived features obtained from experience knowledge to the landslide prediction model to form new features, the processor 502 specifically implements the following steps:

[0093] When the derived features are features of the shape before and after rainfall, the rainfall data corresponding to each slope unit is sorted to obtain a sorting result, and the shape before and after rainfall of the sorting result is determined according to a set index matching type. The features of the shape before and after rainfall within one day, the features of the shape before and after rainfall within one week, and the features of the shape before and after rainfall within one month are added to the landslide prediction model according to time periods.

[0094] In an embodiment, when implementing the step of adding derived features obtained from experience knowledge to the landslide prediction model to form new features, the processor 502 specifically implements the following steps:

[0095] When the derived features are rock mass freeze-thaw index, the rock mass freeze-thaw index is simplified to obtain a processing result, and the processing result is added to the landslide prediction model to obtain new features.

[0096] In an embodiment, when the derived feature is the rock mass freeze-thaw index, the processor 502 performs the following steps to simplify the rock mass freeze-thaw index to obtain the processing result:

[0097] When the derived feature is the rock mass freeze-thaw index, the temperature corresponding to each slope unit in a preset time period is extracted, and the diurnal difference is performed to obtain a temperature difference; it is determined whether the temperature difference is greater than a preset threshold; if the temperature difference is greater than the preset threshold, the rock mass freeze-thaw frequency is calculated, and the ratio of the rock mass freeze-thaw frequency to a relevant constant of the rock mass is taken as a simplified index to obtain the processing result. If the temperature difference is not greater than the preset threshold, the temperature corresponding to each slope unit in a preset time period is extracted, and the diurnal difference is performed to obtain a temperature difference.

[0098] It should be understood that, in the embodiments of the present application, the processor 502 can be a central processing unit (CPU), and the processor 502 can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0099] It can be understood by those skilled in the art that all or part of the processes in the method of the above embodiments can be completed by a computer program instructing related hardware. The computer program includes program instructions, and the computer program can be stored in a storage medium, which is a computer readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the above-mentioned embodiments of the method.

[0100] Therefore, the present application also provides a storage medium. The storage medium can be a computer readable storage medium. The storage medium stores a computer program, wherein the computer program is executed by a processor to make the processor perform the following steps:

[0101] The derived feature obtained from experience knowledge is added to the landslide prediction model to form a new feature; the landslide prediction model is trained using the new feature, so as to use the trained landslide prediction model to perform landslide prediction.

[0102] The derived features include single-variable derived features over historical time and multi-variable interaction derived features.

[0103] The single-variable derived features over historical time include features of pre-and post-rain patterns.

[0104] The multi-variable interaction derived features include rock mass freeze-thaw indexes.

[0105] In an embodiment, when the processor implements the step of adding the derived features from empirical knowledge into the landslide prediction model to form new features by executing the computer program, the processor implements the following steps:

[0106] When the derived features are features of pre-and post-rain patterns, the rainfall data corresponding to each slope unit is sorted to obtain a sorting result; the pre-and post-rain patterns of the sorting result are determined according to a set index matching type; and features of pre-and post-rain patterns within one day, features of pre-and post-rain patterns within one week, and features of pre-and post-rain patterns within one month are added into the landslide prediction model according to time periods.

[0107] In an embodiment, when the processor implements the step of adding the derived features from empirical knowledge into the landslide prediction model to form new features by executing the computer program, the processor implements the following steps:

[0108] When the derived features are rock mass freeze-thaw indexes, the rock mass freeze-thaw indexes are simplified to obtain a processing result; and the processing result is added into the landslide prediction model to obtain new features.

[0109] In an embodiment, when the processor implements the step of, when the derived features are rock mass freeze-thaw indexes, simplifying the rock mass freeze-thaw indexes to obtain a processing result by executing the computer program, the processor implements the following steps:

[0110] When the derived features are rock mass freeze-thaw indexes, the temperature corresponding to each slope unit within a preset time period is extracted and diurnal difference is performed to obtain a temperature difference; it is determined whether the temperature difference is greater than a preset threshold; if the temperature difference is greater than the preset threshold, the number of freeze-thaw of the rock mass is calculated, and the ratio value of the number of freeze-thaw of the rock mass to a relevant constant of the rock mass is taken as a simplified index to obtain the processing result; and if the temperature difference is not greater than the preset threshold, the step of, when the derived features are rock mass freeze-thaw indexes, extracting the temperature corresponding to each slope unit within a preset time period and performing diurnal difference to obtain a temperature difference is executed.

[0111] The storage medium can be a U disk, a mobile hard disk, a read-only memory (ROM), a magnetic disk or an optical disk, and various computer readable storage media that can store program codes.

[0112] Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized in electronic hardware, computer software or a combination of both. In order to clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been described in the above description in a general manner. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0113] In several embodiments provided by the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of each unit is only a logical functional division, and actual implementation can have another division manner. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0114] The steps in the method embodiments of the present application can be adjusted, combined and reduced in sequence according to actual needs. The units in the device embodiments of the present application can be combined, divided and reduced according to actual needs. In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit.

[0115] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a storage medium. Based on such understanding, the technical solutions of the present application essentially or the parts that make contributions to the prior art, or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a terminal or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application.

[0116] The above description is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical range disclosed by the present application, and these modifications or replacements should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for improving the accuracy of landslide prediction, characterized by, The method comprises the following steps: adding a derived feature obtained from experience knowledge into a landslide prediction model to form a new feature; training the landslide prediction model by using the new feature, so as to use the trained landslide prediction model to make landslide prediction; the derived feature comprises a single variable in historical time and a multi-variable interaction derived feature; the single variable in historical time comprises a feature of rainfall before and after the rainfall; the step of adding the derived feature obtained from experience knowledge into the landslide prediction model to form the new feature comprises the following steps: when the derived feature is the feature of rainfall before and after the rainfall, sorting rainfall data corresponding to each slope unit to obtain a sorting result; determining the form of the rainfall before and after the rainfall according to a set index matching type; adding a feature of rainfall form within one day, a feature of rainfall form within one week and a feature of rainfall form within one month into the landslide prediction model according to time periods; the multi-variable interaction derived feature comprises a rock mass freeze-thaw index; the step of adding the derived feature obtained from experience knowledge into the landslide prediction model to form the new feature comprises the following steps: when the derived feature is the rock mass freeze-thaw index, performing a simple processing on the rock mass freeze-thaw index to obtain a processing result; adding the processing result into the landslide prediction model to obtain the new feature; the step of performing the simple processing on the rock mass freeze-thaw index to obtain the processing result when the derived feature is the rock mass freeze-thaw index comprises the following steps: when the derived feature is the rock mass freeze-thaw index, extracting temperature corresponding to each slope unit within a preset time period and performing a day-night difference to obtain a temperature difference; judging whether the temperature difference is greater than a preset threshold value; if the temperature difference is greater than the preset threshold value, calculating a rock mass freeze-thaw frequency and taking a ratio value of the rock mass freeze-thaw frequency and a related constant of the rock mass as a simple index to obtain the processing result; if the temperature difference is not greater than the preset threshold value, performing the step of extracting the temperature corresponding to each slope unit within the preset time period and performing the day-night difference to obtain the temperature difference when the derived feature is the rock mass freeze-thaw index.

2. Landslide prediction accuracy improving apparatus, characterized by, The device uses the landslide prediction accuracy improving method according to claim 1, and comprises: an adding unit configured to add a derived feature obtained from experience knowledge into a landslide prediction model to form a new feature; a training unit configured to train the landslide prediction model by using the new feature, so as to use the trained landslide prediction model to make landslide prediction; the derived feature comprises a single variable in historical time and a multi-variable interaction derived feature; the single variable in historical time comprises a feature of rainfall before and after the rainfall; the adding unit comprises: a sorting subunit configured to, when the derived feature is the feature of rainfall before and after the rainfall, sort rainfall data corresponding to each slope unit to obtain a sorting result; a form determining subunit configured to determine the form of the rainfall before and after the rainfall according to a set index matching type; and a feature adding subunit configured to add a feature of rainfall form within one day, a feature of rainfall form within one week and a feature of rainfall form within one month into the landslide prediction model according to time periods.

3. A computer device, comprising: The computer device comprises a memory and a processor, the memory stores a computer program, and the processor implements the method of claim 1 when executing the computer program.

4. A storage medium, characterized by The storage medium stores a computer program, and the computer program can implement the method of claim 1 when executed by a processor.

Images