Natural disaster detection method, computer equipment and storage medium
By comparing the environmental characteristics constructed by radar data with reference data when no natural disasters occur, the problem of low accuracy in natural disaster detection in the existing technology is solved, and timely and accurate detection of natural disasters is achieved, and it is suitable for all-weather monitoring.
Patent Information
- Application Number
- CN202411143547.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-20
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2044-08-20
AI Technical Summary
The existing natural disaster detection methods have low accuracy and are difficult to detect natural disasters such as mudslides, landslides, rockfalls, etc. in a timely and accurate manner.
By obtaining several frames of radar data in the target area, the current environmental characteristics are constructed, and whether the detection data and reference data when there is no natural disaster are distributed in the same category. If it does not, it is determined that natural disasters occur in the target area and improve detection accuracy.
It realizes timely and accurate detection of natural disasters, is suitable for all-weather and all-day monitoring, and improves the applicability and accuracy of natural disaster detection.
Smart Images

Figure CN118655543B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data processing technology, and in particular to a natural disaster detection method, computer equipment and storage medium. Background Art
[0002] Due to the complex and diverse natural geographical conditions, natural disasters such as debris flow, landslide, rockfall, collapse, avalanche, flood, foreign body invasion, etc. may occur under the influence of natural conditions. Such natural disasters are characterized by suddenness, fast flow, large flow, and strong destructive power, and usually destroy roads and railways, causing huge losses.
[0003] Currently, there are problems with natural disaster detection methods, such as low accuracy. Summary of the invention
[0004] The main technical problem solved by the present application is to provide a natural disaster detection method, computer equipment and storage medium, which can improve the accuracy of natural disaster detection.
[0005] In order to solve the above problems, the first aspect of the present application provides a natural disaster detection method, the method comprising: obtaining a number of frames of radar data collected in a target area during a current detection period; constructing current environmental characteristics based on the several frames of radar data; determining detection data according to the current environmental characteristics; in response to the detection data and reference data not belonging to the same type of distribution data, determining that a natural disaster has occurred in the target area; wherein the reference data is data collected when no natural disaster has occurred in the target area.
[0006] In order to solve the above problems, the second aspect of the present application provides a computer device, which includes a memory and a processor coupled to each other, the memory stores program data, and the processor is used to execute the program data to implement any step of the above natural disaster detection method.
[0007] In order to solve the above-mentioned problem, the third aspect of the present application provides a computer-readable storage medium, which stores program data that can be executed by a processor, and the program data is used to implement any step of the above-mentioned natural disaster detection method.
[0008] The above scheme, this application obtains several frames of radar data collected in the target area during the current detection period, constructs current environmental features based on several frames of radar data, and determines the detection data according to the current environmental features, which can better describe the characteristics of the current environment of the target area. Since the reference data is data collected when no natural disasters occur in the target area, that is, it can represent normal data, it can be determined whether the detection data is abnormal data by judging whether the data collected when no natural disasters occur and the detection data of the current detection period belong to the same type of distribution data, so as to determine whether the detection data is abnormal data, so as to determine whether a natural disaster occurs in the target area. When the detection data and the reference data do not belong to the same type of distribution data, it is determined that a natural disaster occurs in the target area, which can improve the accuracy of natural disaster detection, and can detect natural disasters in a timely manner. The use of radar data to detect natural disasters can also detect natural disasters for a long time, such as around the clock, and improve the applicability of natural disaster detection.
[0009] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] In order to more clearly illustrate the technical solutions in this application, the following briefly introduces the drawings required in the description of the embodiments. Obviously, the drawings described below are only some embodiments of this application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:
[0011] Figure 1 It is a flowchart of the first embodiment of the natural disaster detection method of the present application;
[0012] Figure 2 This is a schematic diagram of an example of an embodiment of point cloud data of the present application;
[0013] Figure 3 This is a schematic diagram of another embodiment of the point cloud data of the present application;
[0014] Figure 4 This application Figure 1 A schematic diagram of a flow chart of an embodiment of step S12;
[0015] Figure 5 It is a schematic diagram of an example of an embodiment of point cloud data grouped in the first direction of the present application;
[0016] Figure 6 It is a schematic diagram of an example of an embodiment of point cloud data grouped in the second direction of the present application;
[0017] Figure 7 This is a schematic diagram of an example of an embodiment of the convex hull area of the present application;
[0018] Figure 8 It is a flow chart of the second embodiment of the natural disaster detection method of the present application;
[0019] Fig. 9 This application Figure 8 A schematic diagram of a flow chart of an embodiment of step S21;
[0020] Fig.10 It is a structural schematic diagram of an embodiment of a computer device of the present application;
[0021] Fig.11 It is a structural schematic diagram of an embodiment of a computer device of the present application;
[0022] Fig.12 It is a structural diagram of an embodiment of a computer-readable storage medium of the present application. DETAILED DESCRIPTION
[0023] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0024] The terms "first" and "second" in this application are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Thus, the features defined as "first" and "second" can explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "multiple" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or devices.
[0025] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.
[0026] The term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the objects associated before and after are in an "or" relationship. In addition, "many" in this article means two or more than two. In addition, the term "at least one" in this article means any combination of at least two of any one or more of a plurality of, for example, including at least one of A, B, and C, can mean including any one or more elements selected from the set consisting of A, B, and C.
[0027] The present application provides the following embodiments, and each embodiment is described in detail below.
[0028] It is understandable that the natural disaster detection method in the present application can be executed by a computer device, and the computer device can be any device with processing capabilities, such as a mobile phone, a computer, a server, etc., and the present application does not limit this. In some possible implementations, the natural disaster detection method in the present application can be implemented by a processor calling program data stored in a memory.
[0029] The present application provides a natural disaster detection method that can improve the accuracy of natural disaster detection. Natural disasters include, for example, debris flow, landslide, rockfall, collapse, avalanche, flood, foreign body invasion and other natural disaster phenomena. The present application takes debris flow as an example for explanation, and the present application does not limit this.
[0030] See also Figure 1 , Figure 1 1 is a flow chart of a first embodiment of a method for detecting natural disasters of the present application. The method may include the following steps:
[0031] S11: Acquire several frames of radar data collected in the target area during the current detection cycle.
[0032] A millimeter wave radar device may be set in the target area, and the target area may be detected by the millimeter wave radar device to obtain several frames of radar data. The target area may be an area where natural disasters need to be detected, for example, an area prone to natural disasters. This application does not limit the target area. Several frames may be 1 frame, 2 frames, 3 frames, or more frames, etc. This application does not limit the specific number of several frames.
[0033] The measurement frequency / measurement period of the millimeter-wave radar device can be set according to the application scenarios such as detection requirements. For example, the measurement period can be set to 0.05 seconds and the measurement frequency can be set to 20HZ. The millimeter-wave radar device has a very strong ability to detect moving objects and can be used to detect slightly moving targets. The millimeter-wave radar device can be used to detect moving targets with small speed changes in the target area all day and all night.
[0034] In some embodiments, a detection cycle can be set to detect natural disasters according to the detection cycle. Because natural disasters such as mudslides are sudden, in order to detect natural disasters as early as possible, it is necessary to collect radar data and give detection judgments in a short time. The time required for event detection of natural disasters set in this application can be 30 seconds, recorded as the detection cycle T. It can take into account the collection of radar data and the detection judgment in a short time, and the amount of data is not too small when the time is short, so as to improve the accuracy and timeliness of detection.
[0035] In some implementations, in each detection cycle, a number of frames of radar data for detection may be determined using a sliding window method, a continuous frame method, an interval frame method, etc. This application does not limit the number of sliding frames in the sliding window method or the number of interval frames in the interval frame method.
[0036] Exemplarily, for example, the millimeter-wave radar device collects 6 frames of radar data, and 3 frames of radar data are used in each detection cycle to detect natural disasters. For the sliding window method, if the number of sliding frames is 2, in the first detection cycle, the radar data of frames 1 to 3 are used, and in the second detection cycle, the radar data of frames 3 to 5 are used. For the continuous frame method, in the first detection cycle, the radar data of frames 1 to 3 are used, and in the second detection cycle, the radar data of frames 4 to 6 are used. For the interval frame method, the number of interval frames is 2, in the first detection cycle, the radar data of frames 1, 3 and 5 are used, and in the second detection cycle, the radar data of frames 2, 4 and 6 are used.
[0037] In some embodiments, the radar data includes point cloud data, or the radar data includes point cloud data and attitude data. The point cloud data may include information such as spatial position, radial speed, radar cross section (RCS), etc. The attitude data may represent the attitude angle of the millimeter wave radar device, such as the attitude angles in three axis directions (P, T, Z). Exemplarily, the millimeter wave radar device is provided with a sensor, and the sensor may be used to collect several frames of attitude data. For example, the sensor includes a gyroscope, which may obtain the attitude angle of the millimeter wave radar device.
[0038] For example, see Figure 2, the acquired point cloud data can be in the form of a data structure. Figure 2 It contains 3 frames of point cloud data, and each frame contains 6 point cloud data. The first row represents the frame number (Time), the number of measurements in the frame (Num), and the measurement time of the frame. The rows after the first row (such as rows 2 to 7) represent specific radar measurement points, that is, point cloud targets. The first column of each radar measurement point is the measurement ID, the second to third columns represent the spatial position (such as the coordinates of the X-axis and Y-axis), and the unit can be meters. The fourth column represents the radar reflection area (in db, intensity unit), and the fifth column is the radial velocity (in m / s).
[0039] Each numerical value of the point cloud data has a clear physical meaning. In order to represent the point cloud data more intuitively, several frames of point cloud data can be represented in the form of a visualization. For example, different colors, forms, etc. can be used in the visualization to represent the radial velocities of different point cloud targets. For example, red dots can be used to represent point cloud targets with radial velocities less than zero, that is, point cloud targets close to the millimeter-wave radar device; green dots can be used to represent point cloud targets with radial velocities greater than zero, that is, point cloud targets far away from the millimeter-wave radar device; black dots can be used to represent point cloud targets with radial velocities equal to zero, that is, stationary point cloud targets.
[0040] For example, see Figure 3 ,exist Figure 3 The center position of the horizontal axis with 0 represents the location of the millimeter wave radar equipment. The units of the horizontal and vertical axes are meters (m). Figure 3 The black solid dots in the figure represent stationary point cloud targets, the hollow circles “o” represent point cloud targets approaching the millimeter-wave radar device, and the asterisks “*” represent point cloud targets moving away from the millimeter-wave radar device.
[0041] S12: construct current environment features based on several frames of radar data.
[0042] For the measured radar data, the current environment features are constructed to describe the current environment features of the target area. The point cloud data and the posture data of the radar data can be analyzed to construct the current environment features.
[0043] Optionally, feature extraction is performed on several frames of radar data, such as extracting features of spatial position, radial velocity, radar reflection area, attitude data, etc., to construct current environment features.
[0044] Optionally, statistical description values of several frames of radar data are obtained, for example, statistical description values of point cloud data are obtained, and then the statistical description values of the point cloud data and the posture data are integrated to construct the current environment features.
[0045] In some embodiments, see Figure 4 , the step S12 of the above embodiment can be further expanded. Based on a plurality of frames of radar data, the current environment features are constructed. This embodiment may include the following steps:
[0046] S121: performing clustering processing on a plurality of frames of radar data to obtain a clustering result.
[0047] In some implementations, before executing step S121, the radar data may be grouped. Specifically, several frames of point cloud data may be grouped to obtain point cloud data grouped in different directions, wherein the point cloud data grouped in a preset direction is used for clustering processing.
[0048] See also Figures 5 and 6 , for example, in Figure 5 , Figure 6 The center position of the horizontal axis with 0 represents the location of the millimeter wave radar device. The units of the horizontal and vertical axes are meters (m). The black solid points in the figure represent stationary point cloud targets. Figure 5 The hollow circle "o" indicates the point cloud target approaching the millimeter-wave radar device. Figure 6 The asterisk "*" indicates a point cloud target that is moving away from the millimeter-wave radar device. Radar data can be grouped into point cloud data grouped in the first direction and point cloud data grouped in the second direction. The first direction can represent the direction of arrival, that is, the point cloud target close to the millimeter-wave radar device. The second direction can represent the direction of departure, that is, the point cloud target moving away from the millimeter-wave radar device. Point cloud data grouped in different directions can be represented by corresponding visualization diagrams.
[0049] In step S121, a clustering method may be used to perform clustering processing on several frames of radar data to obtain a clustering result, wherein the clustering method may be a clustering method with a variable threshold. Since the radar data includes point cloud data and posture data, and the posture data generally rarely changes, during the clustering processing, the posture data may not be clustered, that is, the point cloud data of several frames are clustered to obtain a clustering result.
[0050] Optionally, clustering processing can be performed on the point cloud data of preset direction groups to obtain clustering results. The preset direction grouping can be any one or more of the different direction groups. In the case where the preset direction grouping includes a single direction grouping, for example, the preset direction grouping can be a first direction grouping or a second direction grouping, and the preset direction grouping can be set according to the position of the millimeter wave radar device relative to the target area, the specific geographical conditions of the target area, etc. The present application does not limit the preset direction grouping. In the case where the preset direction grouping includes multiple direction groups, clustering processing is performed on the point cloud data of each direction grouping to obtain clustering results. This method can perform subsequent steps on the obtained clustering results in the subsequent process. In this way, data of more important detection directions of the target area can be obtained. For example, in an environment containing a valley, the preset direction can be the direction towards the valley. This method can improve the accuracy of obtaining detection data.
[0051] In some embodiments, during clustering processing of point cloud data of several frames, point cloud objects that meet preset clustering conditions may be clustered into the same cluster. Point cloud objects belonging to the same cluster in the clustering results obtained by the clustering processing meet the preset clustering conditions.
[0052] Among them, the preset clustering conditions include at least one of the following: the target distance between the spatial positions of any two point cloud targets is less than or equal to the preset distance threshold, the radial velocity difference is less than or equal to the preset velocity difference threshold, the radar reflection area difference is less than or equal to the preset area difference threshold, and the number of point cloud targets included in the cluster is greater than the preset number threshold.
[0053] The radial velocity difference is obtained by using the difference between the radial velocities of two point cloud targets. The radar reflection area difference is obtained by using the difference between the radar reflection areas of two point cloud targets. The target distance is obtained by using the difference between the spatial positions of two point cloud targets.
[0054] In some embodiments, the target distance between any two point cloud targets is calculated using the distance difference between the spatial positions of the two point cloud targets and the ellipse coefficient. Among them, the ellipse coefficient may include a first axis coefficient and a second axis coefficient, and the first axis coefficient and the second axis coefficient may be the same or different. The target distance can be calculated using the ellipse formula using the distance difference between the spatial positions of the two point cloud targets and the ellipse coefficient. Specifically, the difference between the coordinates of the two point cloud targets in the X-axis and Y-axis is obtained, the X-axis difference is used as the X variable of the ellipse formula, and the Y-axis difference is used as the Y variable. Then, the ratio between the square of the X-axis difference and the square of the first axis coefficient and the sum of the ratio between the square of the Y-axis difference and the square of the second axis coefficient are obtained, and the square root of the sum is taken to obtain the target distance. It can be understood that the specific value of the ellipse coefficient can be set according to the specific application scenario, and this application does not limit this.
[0055] In some implementations, the clustering process may be expressed using the following formula:
[0056] , , (1)
[0057] (2)
[0058] (3)
[0059] (4)
[0060] (5)
[0061] in Represents a point cloud target, whose spatial position is , and its radial velocity is , its radar reflection area is . Represents a point cloud target in cluster k or class k, and its spatial position is , and its radial velocity is , its radar reflection area is . Represents the target distance between the spatial positions of two point cloud targets, , They represent the elliptic coefficients respectively, such as represents the first axis coefficient, Represents the second axis coefficient. The target distance can be expressed as the "ellipse" distance between two point cloud targets. Represents the radial velocity difference between two point cloud objects. Represents the difference in radar reflection area between two point cloud targets. Indicates the preset distance threshold, Respectively represent the preset speed difference threshold, Indicates the preset area difference threshold. Representation clustering The number of point cloud objects in . Indicates the preset quantity threshold.
[0062] Among them, the preset speed difference threshold , Radar reflection area difference With point cloud target The above thresholds can be set according to the natural disaster event or target area to be detected, and the elliptic coefficient can be related to the performance of the millimeter wave radar device. This application does not limit the coefficients, thresholds, etc.
[0063] Due to the complex geographical conditions in nature, such as the influence of mud, trees, wind and other objects, clustering of several frames of radar data can reduce errors caused by environmental factors, thereby improving the accuracy of the overall description of environmental characteristics.
[0064] S122: Using the statistical description value of the clustering result, construct the current environment characteristics.
[0065] In the above clustering results, preset statistics can be performed on each target point cloud of the cluster that meets the preset clustering conditions to obtain a statistical description value of the clustering result, and then the statistical description value of the clustering result is used to construct the current environment characteristics.
[0066] In some embodiments, the statistical descriptive value includes at least one of the following: average speed , speed variance , average intensity , intensity variance , convex hull area Since the point cloud data contains spatial position, radial velocity and radar reflection area, the spatial position, radial velocity and radar reflection area can be counted respectively to obtain the statistical description value of the clustering result. Specifically, the mean / median and variance of the radial velocity of each point cloud target in the clustering result can be obtained respectively to obtain the average velocity and speed variance Get the average / median and variance of the radar reflection area of each point cloud target in the clustering result, and get the average intensity and intensity variance Perform convex hull statistics on the spatial position of each point cloud target in the clustering result to obtain the convex hull area It is understandable that the present application may determine the statistical description value according to specific application scenarios and the like, and the statistical description value of the present application is not limited thereto.
[0067] See also Figure 7 For example, the convex hull area can be calculated by a variety of convex hull algorithms. For example, the divide-and-conquer method, Graham scanning method, or Jarvis stepping method can be used. First, the vertices of the convex hull of each point cloud target are determined, and then the area of the polygon defined by the vertices of the convex hull is calculated using a geometric method to obtain the convex hull area.
[0068] In some implementations, the statistical description value may be obtained in the following manner:
[0069] (6)
[0070] (7)
[0071] (8)
[0072] (9)
[0073] (10)
[0074] in, Representation clustering The average speed in Representation clustering The speed variance in Representation clustering The average intensity of Representation clustering The intensity variance in Representation clustering The convex hull area of . represents the average function, represents the variance function, Represents the convex hull area function.
[0075] Optionally, the current environment characteristics can be constructed based on the statistical description values.
[0076] Optionally, during the process of collecting radar data, the posture of the millimeter wave radar device may change. In order to reduce the impact of the posture change, the statistical description value and the posture data can be combined to obtain the current environment characteristics.
[0077] In some embodiments, the current environment features It can be expressed as:
[0078] (11)
[0079] Among them, P, T, and Z represent attitude data, which are the three attitude angles of the millimeter wave radar device.
[0080] The above scheme constructs the current environment features by utilizing the statistical description value of the clustering results, which can better describe the overall situation or overall trend of several frames of radar data. In addition, adding posture data to the current environment features can reduce the impact of posture changes of millimeter-wave radar equipment.
[0081] S13: Determine detection data according to current environmental characteristics.
[0082] According to the current environmental characteristics, detection data can be constructed, wherein the detection data can be obtained by an environmental data model constructed based on the current environmental characteristics.
[0083] In general, the natural geographical environment is safe most of the time, for example, 99% of the time, in a year, that is, no natural disasters occur. Therefore, most of the time, the radar data collected by the millimeter-wave radar equipment is normal / safe data. There is only one factor that can affect the misjudgment: the environmental wind speed. For example, short-term low-speed directional wind, the environmental characteristics of short-term low-speed directional wind (such as statistical description value , , , , ) and the environmental characteristics of natural disasters such as debris flow (such as statistical descriptive values , , , , ) have certain similarities. In special scenarios at certain specific times, it may easily lead to false alarms. Therefore, this application also adds environmental wind speed, and constructs an environmental data model based on environmental characteristics and environmental wind speed, so that the detection data can be determined based on the current environmental characteristics, thereby improving the accuracy of natural disaster detection.
[0084] In some embodiments, after the millimeter wave radar device is set up in the target area, the ambient wind speed at the location of the millimeter wave radar device can be measured, and the ambient wind speed is recorded as Y. The environmental data model can be constructed based on the preset constraint relationship between the measured ambient wind speed and the environmental characteristics. The preset constraint relationship includes the relationship between the ambient wind speed, the environmental characteristics, and the influence coefficient.
[0085] For example, the environmental data model can be constructed based on the relationship between the product of the environmental characteristics and the influence coefficient and the environmental wind speed. For example, the environmental data model is as follows:
[0086] (12)
[0087] (13)
[0088] (14)
[0089] in, Indicates the ambient wind speed, represents the influence coefficient, Indicates environmental characteristics. Indicates intercept or error, etc., influence coefficient Corresponding to each vector element in the environmental characteristics , which can also be called the weight coefficient.
[0090] Then, according to the principle of the constructed environmental data model, the detection data can be constructed based on the current environmental features obtained from the radar data collected in the current detection period. Specifically, a preset constraint relationship between the current environmental features and the environmental wind speed is constructed to obtain the detection data; wherein the preset constraint relationship includes the relationship between the environmental wind speed and the current environmental features and the influence coefficient.
[0091] Exemplarily, the test data and reference data are described below respectively:
[0092] In some embodiments, in most time periods, no natural disasters occur, and reference data (referred to as data set A) can be obtained according to the principle of the constructed environmental data model, wherein the reference data is data collected when no natural disasters occur in the target area. The first environmental wind speed is the environmental wind speed detected by the environment in which the millimeter wave radar device is located when detecting radar data corresponding to the reference environmental feature. The reference data may include a preset constraint relationship between the reference environmental feature and the first environmental wind speed and the first influence coefficient.
[0093] The reference data can be expressed by the following formula:
[0094] (15)
[0095] in, Indicates the first ambient wind speed, represents the first influence coefficient, Indicates the reference environment characteristics. Reference environment characteristics It is obtained based on data collected when no natural disasters occur in the target area.
[0096] In some embodiments, a preset constraint relationship between the current environmental feature and the second environmental wind speed can be constructed to obtain detection data (recorded as data set B); wherein the preset constraint relationship includes a relationship between the second environmental wind speed, the current environmental feature, and the second influence coefficient. The second environmental wind speed is the environmental wind speed obtained by detecting the radar data corresponding to the current environmental feature and detecting the environment in which the millimeter wave radar device is located.
[0097] The test data can be expressed by the following formula:
[0098] (16)
[0099] in, Indicates the second environment wind speed, represents the second influence coefficient, Indicates the current environment characteristics.
[0100] Due to the influence of measurement errors, ambient wind speed, environmental conditions and other factors, the test data may be different from the reference data. For example, the first ambient wind speed is different from the second ambient wind speed, and the current environmental characteristics are different from the reference environmental characteristics. Note:
[0101] (17)
[0102] (18)
[0103] S14: In response to the detection data and the reference data not belonging to the same type of distribution data, determining that a natural disaster has occurred in the target area; wherein the reference data is data collected when no natural disaster has occurred in the target area.
[0104] It is determined whether the detection data and the reference data belong to the same type of distribution data. If they do, in response to the detection data and the reference data not belonging to the same type of distribution data, it is determined that a natural disaster occurs in the target area.
[0105] When the millimeter-wave radar equipment is working, it is difficult to collect data corresponding to natural disasters before natural disasters such as mudslides occur. The millimeter-wave radar equipment collects a large amount of data on non-event occurrences, that is, there can be a large amount of data collected when natural disasters do not occur, and there is a lack of data when natural disasters such as mudslides occur. Therefore, this application determines whether a natural disaster has occurred by determining whether the detection data and the reference data belong to the same type of distribution data. This application uses the elimination method: that is, the reference data is used as a normal sample to determine whether the detection data is a normal sample. If so, no abnormal event has occurred, that is, no natural disaster has occurred; if not, an abnormal event has occurred, that is, a natural disaster has occurred; for example, natural disasters include mudslide events.
[0106] In some embodiments, the same type of distribution data may represent the first ambient wind speed in the detection data and the reference data. , Reference environment characteristics and the second ambient wind speed , Current environment characteristics Belong to the same distribution, or the same constraint relationship. If the test data and the reference data belong to the same type of distribution data, it means that the test data and the reference data correspond to the same event, and the test data corresponds to the detection that no natural disaster has occurred. If the test data and the reference data belong to the same type of distribution data, it means that the test data and the reference data do not correspond to the same event, and the test data corresponds to the detection that a natural disaster has occurred.
[0107] The above scheme, this application obtains several frames of radar data collected in the target area during the current detection period, constructs current environmental features based on several frames of radar data, and determines the detection data according to the current environmental features, which can better describe the characteristics of the current environment of the target area. Since the reference data is data collected when no natural disasters occur in the target area, that is, it can represent normal data, it can be determined whether the detection data is abnormal data by judging whether the data collected when no natural disasters occur and the detection data of the current detection period belong to the same type of distribution data, so as to determine whether the detection data is abnormal data, so as to determine whether a natural disaster occurs in the target area. When the detection data and the reference data do not belong to the same type of distribution data, it is determined that a natural disaster occurs in the target area, which can improve the accuracy of natural disaster detection, and can detect natural disasters in a timely manner. The use of radar data to detect natural disasters can also detect natural disasters for a long time, such as around the clock, and improve the applicability of natural disaster detection.
[0108] In some embodiments, a distribution statistics test method may be used to determine whether the detection data and the reference data belong to the same type of distribution data. The details of this process may refer to the following embodiments.
[0109] See also Figure 8 , Figure 8 1 is a flow chart of a second embodiment of a natural disaster detection method of the present application. The method may include the following steps:
[0110] S21: Perform distribution statistical test on the test data and reference data to obtain statistical test results.
[0111] Optionally, the steps of this embodiment may be performed before the above step S14.
[0112] In step S21, a preset statistical test method may be used to perform a distribution statistical test on the test data and the reference data to obtain a statistical test result. The preset statistical test method may be a significance test method. For example, a distribution statistical test may be performed on the test data and the reference data according to a significance test method of a regression relationship to obtain a statistical test result.
[0113] In some embodiments, see Fig. 9 , step S21 of the above embodiment can be further expanded. Performing distribution statistics test on the detection data and the reference data to obtain the statistical test results, this embodiment may include the following steps:
[0114] S211: Construct a data statistical model for the detection data and reference data.
[0115] A data statistical model may be constructed for the detection data and the reference data according to the linear estimation theory, for example, a linear regression model may be constructed.
[0116] In some embodiments, the data statistical model includes a constraint relationship between the first environmental wind speed, the second environmental wind speed, the reference environmental characteristics, the current environmental characteristics, the corresponding first influence coefficient, the second influence coefficient, and the measurement errors corresponding to the two data.
[0117] The data statistical model can be expressed by the following formula:
[0118] (19)
[0119] in, , They represent measurement errors respectively, and the measurement errors obey Gaussian distribution.
[0120] In addition, the above data statistical model can be expressed as in the form of for Among them, we can make:
[0121] (20)
[0122] (twenty one)
[0123] S212: Determine assumptions based on the data statistical model.
[0124] According to the data statistical model, the hypothesis condition is determined. Among them, the hypothesis condition indicates that the test data and the reference data belong to the same type of distribution data. That is, if data set A and data set B are the same type of data, then the hypothesis condition is established.
[0125] In some embodiments, the assumption includes that the first influence coefficient of the reference data and the second influence coefficient of the detection data are the same. It can be expressed as: .
[0126] Assumptions The following constraints can be met:
[0127] (twenty two)
[0128] Among them, I and -I represent the identity matrix and the negative identity matrix.
[0129] S213: Perform a preset test using the hypothesis conditions to obtain a test statistic as a statistical test result.
[0130] The pre-set test can be performed using the hypothesis conditions. For example, the pre-set test can be a test method based on the F-test, thereby obtaining a test statistic as a statistical test result.
[0131] In some embodiments, the first variance term and the second variance term can be determined using an assumption, wherein the first variance term includes the difference between the estimated variance term corresponding to the reference data and the test data and the second variance term, and the second variance term includes the sum of squares of deviations corresponding to the test data and the sum of squares of deviations corresponding to the reference data. Based on the first variance term and the second variance term, a test statistic is obtained. Specifically, the test statistic can be obtained based on the ratio relationship between the first ratio of the first variance term and the corresponding first degree of freedom and the second ratio of the second variance term and the corresponding second degree of freedom. The first degree of freedom and the second degree of freedom are related to the number of elements of the environmental features or the influence coefficients in the reference data and the test data.
[0132] The test statistic F can be expressed as follows:
[0133] (twenty three)
[0134] in, represents the first variance term, represents the estimated variance term, represents the second variance term, represents the first degree of freedom, represents the second degree of freedom. The second variance term Including the sum of squares of deviations corresponding to the reference data , the sum of squares of deviations corresponding to the test data .
[0135] In some embodiments, based on the assumption that Satisfied constraints, according to the least squares estimation theory with constraints, get the estimated parameters ,as follows:
[0136] (twenty four)
[0137] in, is the first estimated parameter, is the second estimated parameter.
[0138] Based on the above estimated parameters, the first variance term can be obtained, that is, the difference between the estimated variance term corresponding to the reference data and the test data and the second variance term .
[0139] Using the first estimated parameter , we can get the sum of squares of deviations corresponding to the reference data , and, using the second estimated parameter , we can get the sum of squares of deviations corresponding to the detection data Specifically, the ambient wind speed of the reference data can be obtained , the first estimated parameter and reference environment characteristics The square difference between the products of the reference data is obtained by Specifically, the ambient wind speed of the detection data can be obtained , the second estimated parameter and current environmental characteristics The square difference between the products of is used to obtain the sum of squares of deviations corresponding to the detection data. .
[0140] In some embodiments, the sum of squared deviations corresponding to the reference data is It can be expressed as:
[0141] (25)
[0142] The sum of squares of deviations corresponding to the test data It can be expressed as:
[0143] (26)
[0144] In some implementations, the above test statistic F can be converted into another expression to be expressed as follows:
[0145] (27)
[0146] in, represents the first variance term, represents the second variance term, represents the first degree of freedom, represents the second degree of freedom and b is a constant. represents the number of samples of the reference data set A. Indicates the number of samples of the test data, i.e., the data set B. For example, m is 9. for .
[0147] In some embodiments, the test statistic F follows an F distribution For example, the test statistic F follows the F distribution .
[0148] S22: According to the statistical test results, determine whether the test data and the reference data belong to the same type of distribution data.
[0149] According to the statistical test results, it can be determined whether the test data and the reference data belong to the same type of distribution data. Specifically, a preset confidence range or rejection domain range is obtained. For example, the threshold of the confidence range is 0.95 or the threshold of the rejection domain range is 0.05. Then, the statistical test result is obtained by comparing the value of the F distribution with the test statistic F by looking up the table in the F distribution table. If the test statistic F is within the rejection domain range, or the test statistic F is not within the confidence range, it means that the reference data (data set A) and the test data (data set B) are not the same type of distribution data, then the test data is regarded as abnormal data, which means that an abnormal event has occurred, such as a natural disaster.
[0150] In some embodiments, in response to the test statistic meeting the rejection region range, it is determined that the test data and the reference data do not belong to the same type of distribution data, and it is determined that a natural disaster has occurred in the target area. Then, the current collection cycle and recent data are saved, and prompt information is sent to the user or related equipment.
[0151] In some embodiments, if the test statistic F is not within the rejection domain, or the test statistic F is within the confidence range, it means that the reference data (data set A) and the test data (data set B) are the same type of distribution data, and the test data is considered normal data, which means that no abnormal events have occurred, such as no natural disasters. In this case, in response to the test statistic not meeting the rejection domain range, it can be determined that the test data and the reference data belong to the same type of distribution data, and that no natural disasters have occurred in the target area; and / or the test data is added to the reference data.
[0152] In some embodiments, in response to the detection data and the reference data belonging to the same type of distribution data, it is determined that no natural disaster has occurred in the target area, and the detection data is added to the reference data. After being added to the reference data, it can be used as reference data later.
[0153] In some embodiments, the reference data may be updated in each detection cycle, and the updated reference data may be detection data in the previous detection cycle when no natural disaster occurred, or the updated reference data may be reference data obtained by integrating reference data from multiple detection cycles, etc.
[0154] The reference data of multiple detection cycles can be preset and analyzed, and a comprehensive environmental data model describing the target area under normal conditions can be obtained to obtain updated reference data. Among them, the multiple detection cycles can be all detection cycles or the current detection cycle and the detection cycle within the previous preset time period, and this application does not limit this. After each addition of detection data, the reference data of multiple detection cycles can be preset and analyzed to obtain updated reference data. The preset analysis can be a time period analysis, a weather condition analysis, a seasonal condition analysis, a comprehensive statistical analysis, etc., and this application does not limit this. For example, due to the large differences in data sets under different ambient wind speeds, reference data can be uniformly based on ambient wind speeds.
[0155] In some implementations, the process of obtaining reference data can refer to the specific implementation process of obtaining detection data described above, and radar data collected in the most recent period of the current detection cycle can be obtained to construct reference environmental features, thereby modeling and obtaining reference data. The fluctuation range of the reference environmental features in the reference data is statistically analyzed to determine the confidence range or the rejection domain range.
[0156] The above scheme performs distribution statistical tests on the detection data and the reference data to obtain statistical test results. Based on the statistical test results, it is determined whether the detection data and the reference data belong to the same type of distribution data. This can improve the accuracy of testing whether the detection data is abnormal data and whether it belongs to the same distribution data as the reference data. In this way, the accuracy of natural disaster detection can be improved.
[0157] This application only provides technology. The data required for specific applications shall be based on actual conditions. However, users are reminded that when actually using this technology, the collection and processing of data should comply with laws and regulations related to data and user information protection.
[0158] It is understandable that in the above method of the specific implementation mode, the writing order of each step does not mean a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.
[0159] With respect to the above-mentioned embodiments, the present application provides a natural disaster detection device for implementing the above-mentioned natural disaster detection method.
[0160] See also Fig.10 , Fig.10 It is a structural diagram of an embodiment of a computer device of the present application.
[0161] The natural disaster detection device 30 includes: a detection module 31 , a feature module 32 , a data module 33 and a determination module 34 .
[0162] The detection module 31 is used to obtain a number of frames of radar data collected in the target area during the current detection period.
[0163] The feature module 32 is used to construct the current environment features based on a number of frames of radar data.
[0164] The data module 33 is used to determine the detection data according to the current environmental characteristics.
[0165] The determination module 34 is used to determine that a natural disaster has occurred in the target area in response to the detection data and the reference data not belonging to the same type of distribution data; wherein the reference data is data collected when no natural disaster has occurred in the target area.
[0166] It should be noted that the natural disaster detection device provided in the above embodiment and the natural disaster detection method provided in the above embodiment belong to the same concept, wherein the specific manner in which each module and unit performs the operation has been described in detail in the method embodiment and will not be repeated here. In practical applications, the natural disaster detection device provided in the above embodiment can distribute the above functions to different functional modules as needed, that is, divide the internal structure of the device into different functional modules to complete all or part of the functions described above, and this application does not limit this.
[0167] For the above embodiment, the present application provides a computer device, see Fig.11 , Fig.11 4 is a schematic diagram of a computer device of the present application. The computer device 40 includes a memory 41 and a processor 42, wherein the memory 41 and the processor 42 are coupled to each other, the memory 41 stores program data, and the processor 42 is used to execute the program data to implement the steps of any embodiment of the above-mentioned natural disaster detection method.
[0168] In this embodiment, the processor 42 may also be referred to as a CPU (Central Processing Unit). The processor 42 may be an integrated circuit chip having the ability to process signals. The processor 42 may also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gates or transistor logic devices, or discrete hardware components. A general-purpose processor may be a microprocessor or the processor 42 may also be any conventional processor, etc.
[0169] The method of the above embodiment can be implemented in the form of a computer program, so the present application proposes a computer readable storage medium, see Fig.12 , Fig.12 The computer-readable storage medium 50 stores program data 51 that can be run by a processor, and the program data 51 can be executed by the processor to implement the steps of any embodiment of the above-mentioned natural disaster detection method.
[0170] The computer-readable storage medium 50 of this embodiment may be a medium that can store the program data 51, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or may be a server storing the program data 51. The server may send the stored program data 51 to other devices for execution, or may execute the stored program data 51 by itself.
[0171] In some embodiments, the functions or modules included in the device provided in the above embodiments of the present application can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments. For the sake of brevity, the present application will not go into details here.
[0172] The above description of the various embodiments tends to emphasize the differences between the various embodiments. The same or similar aspects can be referenced to each other. For the sake of brevity, this application will not go into details here.
[0173] The above descriptions are merely embodiments of the present application and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.
Claims
1. A method for detecting natural disasters, characterized in that: include: Obtain several frames of radar data collected in the target area during the current detection period; Based on the plurality of frames of radar data, construct current environment features; Determining detection data according to the current environmental characteristics; In response to the detection data and the reference data not belonging to the same type of distribution data, it is determined that a natural disaster has occurred in the target area; wherein the reference data is data collected when no natural disaster has occurred in the target area; the reference data includes a preset constraint relationship between a reference environmental feature and a first environmental wind speed and a first influence coefficient, wherein the first environmental wind speed is an environmental wind speed detected by detecting an environment in which a millimeter-wave radar device is located when detecting radar data corresponding to the reference environmental feature; Among them, determining the detection data based on the current environmental characteristics includes: constructing a preset constraint relationship between the current environmental characteristics and the second environmental wind speed to obtain the detection data; wherein the detection data includes a preset constraint relationship between the current environmental characteristics and the second environmental wind speed and a second influence coefficient, and the second environmental wind speed is the environmental wind speed obtained by detecting the environment in which the millimeter wave radar device is located when detecting the radar data corresponding to the current environmental characteristics.
2. The method according to claim 1, characterized in that The constructing the current environment feature based on the plurality of frames of radar data includes: Performing clustering processing on the plurality of frames of radar data to obtain a clustering result; The current environment characteristics are constructed using the statistical description values of the clustering results.
3. The method according to claim 2, characterized in that The radar data includes point cloud data; the clustering process is performed on the plurality of frames of radar data to obtain a clustering result, including: Performing clustering processing on the point cloud data of the plurality of frames to obtain the clustering result; Among them, the point cloud targets belonging to the same cluster in the clustering results meet the preset clustering conditions, and the preset clustering conditions include at least one of the following: the target distance between the spatial positions of any two point cloud targets is less than or equal to the preset distance threshold, the radial velocity difference is less than or equal to the preset velocity difference threshold, the radar reflection area difference is less than or equal to the preset area difference threshold, and the number of point cloud targets contained in the cluster is greater than the preset number threshold.
4. The method according to claim 3, characterized in that The target distance between any two point cloud targets is calculated using the distance difference between the spatial positions of the two point cloud targets and the ellipse coefficient; And / or, the clustering of the point cloud data of the plurality of frames to obtain the clustering result includes: The point cloud data of the plurality of frames are grouped to obtain point cloud data grouped in different directions, wherein the point cloud data grouped in a preset direction is used for clustering processing.
5. The method according to claim 2, characterized in that: The radar data includes point cloud data and posture data; the current environment features are constructed by using the statistical description value of the clustering result, including: Obtaining a statistical description value of the clustering result; wherein the statistical description value includes at least one of the following: average speed, speed variance, average intensity, intensity variance, and convex hull area; The current environment characteristics are obtained by combining the statistical description value and the posture data.
6. The method according to claim 1, characterized in that In response to the detection data and the reference data not belonging to the same type of distribution data, the method further comprises: Performing distribution statistics test on the detection data and the reference data to obtain statistical test results; According to the statistical test result, it is determined whether the detection data and the reference data belong to the same type of distribution data.
7. The method according to claim 6, characterized in that The performing of distribution statistical test on the detection data and the reference data to obtain statistical test results includes: Constructing a data statistical model for the detection data and reference data; Determining assumptions based on the data statistical model; The assumptions are used to perform a preset test to obtain a test statistic, which is used as the statistical test result.
8. The method according to claim 7, characterized in that The assumption condition indicates that the detection data and the reference data belong to the same type of distribution data; And / or, performing a preset test using the hypothesis condition to obtain a test statistic includes: Determine a first variance term and a second variance term by using the assumption, wherein the first variance term includes a difference between an estimated variance term corresponding to the reference data and the detection data and the second variance term, and the second variance term includes a sum of squares of deviations corresponding to the detection data and a sum of squares of deviations corresponding to the reference data; Obtaining the test statistic based on the first variance term and the second variance term; And / or, judging whether the detection data and the reference data belong to the same type of distribution data according to the statistical test result, comprises: In response to the test statistic being consistent with the rejection region range, it is determined that the detection data and the reference data do not belong to the same type of distribution data.
9. The method according to claim 1, characterized in that: Also includes: In response to the detection data and the reference data belonging to the same type of distribution data, determining that no natural disaster has occurred in the target area, and / or, adding the detection data to the reference data; And / or, performing a preset analysis on the reference data of multiple detection cycles to obtain the updated reference data.
10. A computer device, characterized in that: The invention comprises a memory and a processor coupled to each other, wherein the memory stores program data, and the processor is used to execute the program data to implement the steps of the method according to any one of claims 1 to 9.
11. A computer-readable storage medium, characterized in that: Program data that can be executed by a processor are stored, and the program data is used to implement the steps of the method according to any one of claims 1 to 9.
Citation Information
Patent Citations
Intelligent detection and evaluation method for natural disaster risk and application
CN111027402A
Method and device for determining deviation of landslide mass, electronic equipment and storage medium
CN116299533A
Power transmission line icing thickness detection and ice melting disaster prevention method and system based on laser point cloud
CN116428992A
Mountain torrent disaster detection and early warning equipment
CN218729330U
Method and device for measuring flow velocity of fluid surface, and storage medium
WO2023071909A1
