Water immersion early warning method and related equipment thereof

Abstract

The invention provides a water immersion early warning method and related equipment thereof. The water immersion early warning method comprises the following steps: determining environmental data in a target area and state data of corresponding monitored target equipment; determining water immersion risk data based on the environment data and the state data; performing anomaly identification on the water immersion risk data through a preset anomaly algorithm to obtain an anomaly identification result; and executing a corresponding water immersion early warning scheme based on the exception recognition result in combination with a preset exception handling strategy. Through the steps of the method, fusion and intelligent risk identification of multi-source monitoring data can be realized, potential water immersion hidden dangers can be found in time, accurate early warning is realized through a hierarchical response mechanism, and the reliability and response efficiency of a water immersion monitoring system are improved.

Description

Technical Field

[0001] This invention relates to the field of substation fire monitoring technology, and in particular to a water immersion early warning method, device, electronic equipment and its storage medium. Background Technology

[0002] With the continuous expansion of urban infrastructure, the safety issues of waterproofing and drainage in enclosed spaces such as subway tunnels, underground parking garages, and substations are becoming increasingly prominent. Traditional water immersion monitoring methods mainly rely on water level sensors or manual inspections. When rainfall, poor drainage, or pipe network damage occurs, localized water accumulation or even equipment damage can easily occur, posing significant safety hazards. To ensure the safety of key areas, the industry is gradually introducing intelligent solutions that incorporate multi-sensor data acquisition, networked monitoring, and remote early warning systems to improve the real-time performance and reliability of monitoring.

[0003] Existing flood warning methods still suffer from several drawbacks. They rely solely on a single sensor measurement point, lack comprehensive analysis of environmental and equipment status data, and fail to accurately reflect overall risk. They often use static threshold judgments, failing to consider historical fluctuations and environmental changes, which can easily lead to false alarms or missed alarms. Most methods employ simple threshold triggering algorithms, making it difficult to identify multiple types of abnormal events and different severity levels. After an alarm is triggered, there is a lack of adaptive response mechanisms based on the anomaly level, resulting in delayed strategy adjustments. In long-term monitoring environments, existing equipment often operates continuously at high power consumption, failing to distinguish between monitoring cycles and static standby phases, leading to low energy efficiency and high maintenance costs.

[0004] Therefore, existing flood warning methods have the problem of failing to achieve multi-source data fusion, dynamic power consumption control, and intelligent early warning response based on anomaly level. Summary of the Invention

[0005] This invention provides a water immersion early warning method to solve the problems of existing water immersion early warning methods, which cannot achieve multi-source data fusion, dynamic power consumption control, and intelligent early warning response based on anomaly level.

[0006] In a first aspect, the present invention provides a method for early warning of water immersion, the method comprising the following steps: Determine the environmental data in the target area and the status data of the corresponding monitored target devices; Based on the environmental and status data, water immersion risk data is determined; Anomaly identification results are obtained by using a preset anomaly algorithm to identify anomalies in the water immersion risk data. Based on the anomaly identification results and in conjunction with the preset anomaly handling strategy, the corresponding water immersion early warning plan is executed.

[0007] Optionally, the environmental data includes water level data, water immersion depth data, and humidity data; the status data includes operating voltage, communication connection status, and tilt angle data; and the environmental data for determining the target area and the status data of the corresponding monitored target equipment include: By using preset environmental sensors, environmental data in the target area is collected in real time to obtain water level data, water immersion depth data, and humidity data; The device detection module detects the status data of the target device to obtain the corresponding operating voltage, communication connection status and tilt angle data. By collecting video stream data from the target device and verifying the authenticity of the water level data, immersion depth data, humidity data, operating voltage, communication connection status, and tilt angle data, the environmental data in the target area and the status data of the corresponding monitored target device are determined.

[0008] Optionally, determining the flood risk data based on the environmental data and status data includes: Feature extraction is performed on the environmental data and state data to obtain corresponding environmental feature vectors and state feature vectors; The environmental feature vector and the state feature vector are fused together to obtain the feature fusion vector. Based on the feature fusion vector, key feature parameters within the target area are determined. These key feature parameters include water level change rate, immersion duration, tilt angle change trend, and humidity change amplitude data. The key feature parameters are compared with historical key feature parameter thresholds to generate a water immersion risk index; Based on the water immersion risk index, water immersion risk data within the target area is determined.

[0009] Optionally, before comparing the key feature parameters with historical key feature parameter thresholds to generate a water immersion risk index, the method further includes: Determine the historical monitoring data and corresponding trend data of water immersion risk within the target area; Based on the historical monitoring data and the corresponding water immersion risk change trend data, the upper and lower limits of the thresholds of the corresponding key feature parameters are adjusted to obtain the target threshold. Based on the target threshold, the historical key feature parameter thresholds corresponding to the current key feature parameters are updated and replaced.

[0010] Optionally, the step of using a preset anomaly algorithm to identify anomalies in the water immersion risk data and obtaining anomaly identification results includes: Based on the water immersion risk data, a corresponding multidimensional feature matrix is ​​constructed; The deviation of the multidimensional feature matrix is ​​calculated to obtain the difference value between each feature and the normal data; When the difference value exceeds the preset difference threshold range of the corresponding feature, the corresponding anomaly type is determined, and the anomaly level is classified according to the size of the difference value. An anomaly identification result is generated based on the anomaly type and anomaly level.

[0011] Optionally, when the difference value exceeds a preset difference threshold range for the corresponding feature, determining the corresponding anomaly type and classifying the anomaly level according to the magnitude of the difference value includes: For differences exceeding a preset difference threshold range for the corresponding feature, trace the source to determine the corresponding anomaly type; Based on the anomaly type, extract the associated abnormal change feature data from the water immersion risk data; Based on the range of difference values ​​corresponding to the abnormal change characteristic data, the abnormality level is divided into a prompt level, a warning level, and an alarm level.

[0012] Optionally, the step of executing a corresponding water immersion early warning scheme based on the anomaly identification result and in conjunction with a preset anomaly handling strategy includes: Based on the anomaly type and anomaly level corresponding to the anomaly identification results, a preset anomaly handling strategy is matched; Based on the matching anomaly handling strategy, a corresponding early warning instruction is generated and executed. The early warning instruction is used to control the alarm device, remote push terminal or video surveillance module of the target area to execute the early warning response. After completing the early warning response, record the abnormal event information and update or optimize the abnormal handling strategy.

[0013] Secondly, the present invention also provides a flood warning device, the flood warning device comprising: The first determining module is used to determine the environmental data in the target area and the status data of the corresponding monitored target devices. The second determining module is used to determine water immersion risk data based on the environmental data and the status data; The first identification module is used to identify anomalies in the water immersion risk data using a preset anomaly algorithm, and obtain anomaly identification results. The early warning module is used to execute the corresponding water immersion early warning scheme based on the anomaly identification results and in combination with the preset anomaly handling strategy.

[0014] Thirdly, the present invention provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps in the water immersion warning method provided by the present invention.

[0015] Fourthly, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the water immersion early warning method provided by the invention.

[0016] This invention determines environmental data and status data of corresponding monitored target devices within a target area; based on the environmental and status data, it determines water immersion risk data; through a preset anomaly algorithm, it identifies anomalies in the water immersion risk data to obtain anomaly identification results; based on the anomaly identification results and combined with a preset anomaly handling strategy, it executes a corresponding water immersion early warning scheme. Through the above method steps, it is possible to achieve the fusion of multi-source monitoring data and intelligent risk identification, promptly detect potential water immersion hazards, and achieve accurate early warning through a graded response mechanism, thereby improving the reliability and response efficiency of the water immersion monitoring system. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a water immersion early warning method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of another water immersion early warning device provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] like Figure 1 As shown, Figure 1This is a flowchart of a flood warning method provided in an embodiment of the present invention. The flood warning method includes the following steps: 101. Determine the environmental data in the target area and the status data of the corresponding monitored target devices.

[0021] In this embodiment of the invention, the above-mentioned flood warning method can be applied to a flood warning platform. The flood warning platform has functions such as flood data processing, flood data transmission and reception, and flood data memory storage. It can be built based on a server or server cluster. The server or server cluster can be an electronic device with flood data processing capabilities.

[0022] The aforementioned flood warning platform adopts a multi-layered distributed architecture design to achieve collaborative operation of data collection, analysis, decision-making, and display. It includes, but is not limited to, a five-layer architecture consisting of a device layer, a driver layer, a service layer, an algorithm layer, and an application layer. Each layer interacts with the other through standardized communication interfaces to form a stable and scalable monitoring and early warning system.

[0023] The equipment layer consists of various monitoring terminals, including water level sensors, immersion sensors, tilt sensors, and on-site cameras. This layer is responsible for the real-time acquisition of environmental parameters and equipment attitude data, and serves as the data foundation layer for the entire platform.

[0024] The driver layer enables inter-device communication and data conversion through multi-protocol compatible hardware driver modules, supporting interfaces such as RS-485, GMAC, and 4G modules. The driver layer also integrates an encryption chip and a watchdog mechanism to ensure data transmission security and the stable operation of the aforementioned flood warning platform. In the event of communication interruption or process abnormality, the driver layer can automatically restart and recover, preventing data loss.

[0025] The service layer is responsible for the underlying service management and data communication of the aforementioned flood warning platform. Its core components include: MQTT communication service: enabling lightweight data transmission between devices and the platform; Web service interface: allowing users to access the platform's functions via a web page; and data processing service: responsible for data parsing, storage, and forwarding, providing structured input for upper-layer algorithms. The service layer constructs a secure and scalable IoT communication foundation.

[0026] The algorithm layer is the intelligent core of the aforementioned flood warning platform, comprising the following sub-layers: a data acquisition sub-layer (performing data verification and anomaly filtering); a risk assessment sub-layer (calculating flood risk levels using feature fusion models); a warning and decision support sub-layer (generating warning strategies by combining historical trend data); and a management and maintenance sub-layer (used for equipment status diagnosis and remote maintenance control). The algorithm layer uses machine learning and statistical analysis techniques to provide accurate risk analysis results to the upper layers.

[0027] The application layer is the user interaction layer, including modules such as "overall view, intelligent early warning, smart reports, and equipment status monitoring". Users can view real-time monitoring data and alarm status on the overview interface. The aforementioned water flooding early warning platform automatically generates reports and trend analysis charts, and supports event query, download, and export operations to achieve data-driven decision support.

[0028] The target area mentioned above can be a spatial range that needs to be monitored for water immersion safety, such as subway tunnels, underground garages, pump rooms, cable trenches, power distribution rooms, or the area at the bottom of machine rooms. In this embodiment, it can be the monitoring area of ​​a fire substation.

[0029] Specifically, the monitoring area can be divided, with each zone corresponding to a unique number. Through RS-485 or Ethernet communication, the sensor nodes corresponding to each ZoneID can be identified in real time, thereby determining the spatial boundaries of the target area, the sensor placement locations, and the data acquisition path. During this determination process, the aforementioned flood warning platform can also combine geographic information data (GIS coordinates) with on-site equipment numbers to achieve a one-to-one mapping between areas and equipment, providing spatial basis for subsequent risk location.

[0030] The aforementioned environmental data can be monitoring parameters used to reflect changes in the physical environment within the target area, including but not limited to: water level data: collected by water level sensors deployed at low-lying points or sump pits, representing the real-time change of the water surface relative to the reference height; water immersion depth data: detected by electrode-type or ultrasonic water immersion probes, reflecting the thickness of accumulated water in the area, used to determine the degree of ground water accumulation; humidity data: collected by air humidity sensors or infrared humidity detection modules, reflecting the trend of air humidity changes, used to assist in determining the risk of seepage and dampness.

[0031] Understandably, the aforementioned flood warning platform can periodically sample data from sensors and upload the data via the Modbus protocol. This allows for time synchronization, anomaly filtering, and data integrity verification of the collected data, ultimately determining the current environmental data set. During this determination process, the platform can also automatically eliminate interference points or lost communication packets based on the sampling timestamp and signal strength, ensuring the authenticity and continuity of the environmental data.

[0032] The aforementioned target equipment may be terminal devices deployed in the target area that have environmental monitoring or security protection functions, including but not limited to: environmental monitoring terminals (integrated multi-parameter sensing modules), auxiliary communication terminals (LoRa or 4G gateway nodes), security cameras, lighting and drainage pump control units, etc.

[0033] When identifying target devices, the aforementioned flood warning platform can identify the unique identifier (DeviceID) of each terminal through the device registry and device communication interface, and extract its model, installation location and operating parameters from the device management database.

[0034] Through polling and status feedback mechanisms, the aforementioned flood warning platform can confirm the online status and operating mode of each target device in real time, thereby determining the complete set of monitored objects.

[0035] The aforementioned status data can be a real-time data set used to reflect the working status and communication performance of the target device, including but not limited to: operating voltage: sampled in real time by the device's internal power detection module to reflect the device's power supply stability; communication connection status: determined by heartbeat packet response and network quality indicators (RSSI, latency time) to determine whether the device is online; tilt angle data: collected by tilt sensors installed on the device body to reflect the device's attitude changes under stress or immersion in water, used to determine structural stability.

[0036] Understandably, the aforementioned status data is uniformly encapsulated by the edge acquisition gateway and uploaded to the server of the aforementioned flood warning platform. During the determination process, the aforementioned flood warning platform performs outlier removal, dimension normalization, and data labeling operations to generate a standardized status dataset that can be analyzed.

[0037] In one possible embodiment, the aforementioned flood warning platform dynamically determines the target area data by jointly verifying the data collected by environmental sensors and the operating data of the target equipment. Specifically, it can obtain multi-dimensional monitoring data from environmental sensor nodes; obtain operating status and attitude data from the target equipment; compare the timestamps and spatial consistency of the collected data based on the video stream verification module; and form a unified data frame structure with the verified environmental data and status data, and store it in the data cache layer as input for subsequent flood risk analysis.

[0038] 102. Based on environmental and status data, determine the risk of flooding.

[0039] In this embodiment of the invention, the aforementioned flood risk data can refer to a set of quantitative information used to characterize the current or potential flood risk level of a target area, which can be obtained through comprehensive calculation of environmental monitoring parameters and equipment operating status. It can be stored in the form of a risk index, with a value ranging from 0 to 1; a higher value indicates a higher risk level.

[0040] In this embodiment, the aforementioned flood warning platform can receive environmental datasets and status datasets uploaded by the data acquisition layer, including environmental datasets: water level height, flood depth, and air humidity; and status datasets: equipment operating voltage, communication connection status, and tilt angle.

[0041] The algorithm layer synchronizes and corrects the above data according to timestamps to eliminate timing deviations caused by sampling delays and communication jitter.

[0042] After cleaning and normalization, the environmental and state data are uniformly encapsulated into an input matrix D=[E,S], where E represents the environmental parameter vector and S represents the device state parameter vector.

[0043] By performing hierarchical processing on the input matrix D, feature extraction is achieved. Specifically, derivative operations are performed on environmental features to extract the rate of water level change, the magnitude of humidity change, and the duration of water accumulation; time series analysis is performed on state features to extract the trend of tilt angle change and the stability of operating voltage; and a weighted fusion function is used to extract these features. A comprehensive feature vector F is calculated, where the weights w1 and w2 are automatically adjusted by the aforementioned flood warning platform based on sensor reliability and historical correlation. The fusion result serves as the risk feature vector for the target area, used to further calculate flood risk indicators.

[0044] In one possible embodiment, the aforementioned flood warning platform uses a risk assessment subsystem to comprehensively analyze environmental data and equipment status data to determine the flood risk data for the target area.

[0045] The system receives multi-source input data from sensors and equipment detection modules, including water level, immersion depth, humidity, voltage, communication status, and tilt angle. The risk assessment subsystem performs time synchronization and anomaly filtering to eliminate sampling errors and communication delays.

[0046] The feature extraction module extracts features such as water level rise rate, humidity increase, equipment posture change trend, and voltage fluctuation from continuous data. After standardization, these features are input into the fusion calculation module to generate comprehensive feature results according to weights.

[0047] The fusion results are compared with historical data to calculate the degree of deviation, forming risk indicators that reflect the trend of water immersion and the extent of equipment impact. The risk assessment subsystem outputs quantified risk level data as water immersion risk data.

[0048] During operation, the risk assessment subsystem dynamically adjusts the weights and threshold ranges based on environmental changes, enabling risk data to reflect the on-site working conditions in real time, completing the quantitative assessment of water immersion risk in the target area, and providing basic data for anomaly identification and graded early warning.

[0049] In another possible embodiment, the aforementioned flood warning platform uses the risk assessment subsystem of the algorithm layer to perform feature extraction, feature fusion, threshold comparison and weight calculation to achieve risk quantification output, forming risk indicator data for subsequent anomaly identification, namely flood risk data.

[0050] Through the above methods and steps, the risk index can be dynamically calculated based on the fusion of multi-source data, reflecting changes in water level, humidity fluctuations, and the degree of impact on equipment. It has the ability to adaptively correct thresholds and perform feature weighted fusion, which can reduce false alarms and missed alarms and improve identification accuracy, thereby enabling real-time judgment and hierarchical management of water immersion trends, and providing a reliable basis for subsequent anomaly identification and early warning response.

[0051] 103. By using a preset anomaly algorithm, anomaly identification is performed on the water immersion risk data to obtain anomaly identification results.

[0052] In this embodiment of the invention, the aforementioned preset anomaly algorithm can be a pre-configured anomaly detection model in the algorithm layer, used to identify abnormal trends in risk data. It can perform statistical deviation calculations and time-series difference analysis using a risk feature matrix to comprehensively determine the abnormal correlation between multiple parameters such as water level, humidity, voltage, and tilt angle. Furthermore, it incorporates a built-in difference calculation, clustering comparison, and adaptive threshold adjustment mechanism, which can automatically set identification boundaries based on the historical characteristics of different monitoring points. When the algorithm detects that the change pattern of the input data significantly deviates from the normal operating distribution, it outputs the anomaly type and level.

[0053] Specifically, when flood risk data is input, the deviation of each parameter is calculated sequentially after multi-dimensional feature analysis to identify situations such as sudden changes, abnormal rises in water level, sharp increases in humidity, or equipment tilting exceeding limits. During the anomaly identification phase, the aforementioned flood warning platform compares the change curves of each parameter to determine whether it is a short-term fluctuation or a continuous anomaly, and combines this with historical trends to determine the persistence and severity of the anomaly.

[0054] The above anomaly identification results can be judgment results based on algorithm calculations, used to reflect the actual abnormal state of the target area, including but not limited to anomaly types such as sensor anomalies, water level anomalies, power supply anomalies, etc., anomaly levels such as prompts, warnings, alarms, etc., occurrence time, involved equipment and corresponding data.

[0055] In one possible embodiment, after receiving real-time risk data, the aforementioned flood warning platform calls a preset anomaly algorithm to generate a feature matrix, calculates the difference between each feature value and the normal baseline, and if the difference exceeds the corresponding threshold, triggers the anomaly identification process. Based on the difference magnitude, duration, and feature combination, the anomaly type and level are determined, forming a structured anomaly identification result, which is then synchronized to the intelligent warning module to execute the processing strategy.

[0056] 104. Based on the anomaly identification results and combined with the preset anomaly handling strategy, execute the corresponding water immersion early warning plan.

[0057] In this embodiment of the invention, the aforementioned preset anomaly handling strategy can be a set of response rules pre-established during the deployment phase according to different anomaly types and risk levels, used to guide the execution of early warning responses, including but not limited to a response level mapping table: different handling actions correspond to the level (prompt, warning, alarm) in the anomaly identification results; response object configuration: defining alarm devices (audio-visual alarms, control relays, cameras), notification terminals (management platform, mobile APP), and remote push paths; linkage rules: when a high-level anomaly is detected, multi-channel alarms and video linkage retrospective are triggered simultaneously; strategy update mechanism: combining historical handling effects and false alarm records, the triggering conditions are automatically corrected or optimized.

[0058] In one possible embodiment, the aforementioned flood warning platform generates specific control instructions based on the matched preset strategy and sends them to the relevant execution modules. Specifically, when the aforementioned flood warning platform receives the anomaly identification result, it automatically calls the strategy matching module, reads the processing rules corresponding to the anomaly type and level, and generates a set of warning instructions including "activation of sound and light alarm", "video monitoring linkage", and "remote message push".

[0059] The aforementioned flood warning platform sends instructions to field controllers or cloud servers via a communication bus or IoT protocol, enabling automatic responses from field equipment. Simultaneously, the platform records execution logs, including execution time, action type, execution result, and feedback status, for subsequent traceability and strategy optimization.

[0060] The aforementioned flood warning scheme can be a comprehensive response process executed under abnormal conditions, used to achieve end-to-end protection from detection to handling, including but not limited to on-site warning: activating audible and visual alarm devices to alert on-site personnel to potential risks; remote alarm: pushing warning information to management terminals via SMS, WeChat, and IoT platforms; video linkage: calling corresponding monitoring cameras to capture and stream images in real time to assist manual verification; data recording: saving monitoring data, alarm levels, and processing results of the current event to form an event log; and strategy closed loop: adjusting corresponding thresholds and response strategies based on processing feedback to optimize subsequent warning effects.

[0061] The above methods and steps ensure rapid response and multi-level coordination under different risk levels. The execution process ensures the real-time nature and traceability of the response, thereby improving the safety control capabilities and management efficiency of the aforementioned flood warning platform.

[0062] In this embodiment of the invention, environmental data and status data of the corresponding monitored target devices in the target area are determined; based on the environmental data and status data, water immersion risk data is determined; anomaly identification is performed on the water immersion risk data using a preset anomaly algorithm to obtain anomaly identification results; based on the anomaly identification results and combined with a preset anomaly handling strategy, a corresponding water immersion early warning scheme is executed. Through the above method steps, the fusion of multi-source monitoring data and intelligent risk identification can be achieved, potential water immersion hazards can be detected in a timely manner, and accurate early warning can be achieved through a hierarchical response mechanism, thus improving the reliability and response efficiency of the water immersion monitoring system.

[0063] Optionally, in the steps of determining the environmental data in the target area and the status data of the corresponding monitored target equipment, the environmental data in the target area can be collected in real time by using preset environmental sensors to obtain water level data, water immersion depth data, and humidity data; the status data of the target equipment can be detected by the equipment detection module to obtain the corresponding operating voltage, communication connection status, and tilt angle data; and the environmental data in the target area and the status data of the corresponding monitored target equipment can be determined by collecting video stream data of the target equipment and verifying the authenticity of the water level data, water immersion depth data, humidity data, operating voltage, communication connection status, and tilt angle data.

[0064] In this embodiment of the invention, the aforementioned preset environmental sensor can refer to a sensing component that is deployed in the target area and has real-time sampling and communication functions, specifically including a water level sensor, a water immersion sensor and a humidity sensor, which can be connected to the host through an RS485 or 4G communication interface to realize real-time acquisition of water level, water immersion depth and air humidity.

[0065] The aforementioned device detection module can be a status monitoring unit embedded inside the target device, used to detect the device's power supply, communication, and attitude status in real time. It can periodically sample data through the internal microcontroller unit (MCU) and transmit the data to the host platform via Ethernet or serial port.

[0066] The aforementioned environmental data may include, but is not limited to, water level data, water immersion depth data, and humidity data. Water level data reflects the height of the accumulated water surface and is obtained from ultrasonic or pressure-type water level sensors, used to determine the rate of water rise and fluctuation trends within the area. Water immersion depth data is acquired by electrode-type or float-type water immersion probes, characterizing the depth of water cover on the ground surface and used to determine the level of seepage or flooding. Humidity data is collected by air humidity sensors, reflecting changes in relative humidity and assisting in identifying potential leakage or condensation risks. All three types of data can be uploaded to the aforementioned water immersion early warning platform via a communication driver layer (such as RS485 or MQTT protocol), and after timestamping synchronization, enter the data processing module.

[0067] The aforementioned status data may include, but is not limited to, operating voltage, communication connection status, and tilt angle data. Among them, operating voltage reflects whether the power supply to the device is stable. When the voltage is lower than the set threshold, the aforementioned water immersion warning platform marks a potential power supply anomaly. Communication connection status is determined by the network monitoring module through heartbeat detection to determine whether the device is online. When a disconnection occurs or the delay exceeds the set value, it is automatically reported. Tilting angle data is collected by tilt sensors (such as triaxial MEMS gravity sensors) to identify abnormal posture caused by structural displacement or external impact.

[0068] The aforementioned video stream data can be collected by video surveillance equipment deployed in the target area, providing on-site footage as supplementary verification evidence. It can be uploaded via the network driver layer (GMAC or 4G module), enabling real-time preview and data frame comparison. Understandably, the video footage can be used to determine the presence of obvious physical phenomena such as water accumulation, rising water levels, or equipment tilting, thereby enhancing data credibility.

[0069] In one possible embodiment, the aforementioned flood warning platform confirms the authenticity of the collected data by cross-validating the sensor data and the video stream. Specifically, multi-source consistency verification is performed during the data processing stage. When the water level sensor shows an increase but the video footage does not show any change in water accumulation, the aforementioned flood warning platform automatically marks the data as suspicious and triggers resampling.

[0070] Meanwhile, the aforementioned flood warning platform compares the numerical differences between different sensors. If the humidity continues to rise but the water level and flood depth do not change significantly, the platform will use a rule engine to perform logical checks to prevent misjudgments.

[0071] Optionally, the step of determining flood risk data based on environmental and state data further includes: extracting features from the environmental and state data to obtain corresponding environmental and state feature vectors; fusing the environmental and state feature vectors to obtain a feature fusion vector; determining key feature parameters within the target area based on the feature fusion vector; comparing the key feature parameters with historical key feature parameter thresholds to generate flood risk indicators; and determining flood risk data within the target area based on the flood risk indicators.

[0072] In this embodiment of the invention, the collected environmental data and equipment status data can be statistically analyzed, and trends and rates of change can be analyzed. Specifically, the raw data such as water level, humidity, voltage, and tilt angle can be subjected to sliding sampling and normalization to extract key trends, such as rate of increase, periodic fluctuations and abnormal peaks, in order to reflect the dynamic change patterns of the environment and equipment.

[0073] The aforementioned environmental feature vector can be composed of environmental parameters such as water level, immersion depth, and humidity, used to characterize the external environmental state. In this embodiment, the change, average value, and maximum change amplitude of each parameter in the time series can be mapped into a multi-dimensional vector for fusion analysis with equipment status features.

[0074] The aforementioned state feature vector can be composed of data such as equipment operating voltage, communication connection status, and tilt angle, which are used to reflect the equipment's working stability and structural attitude. By quantifying the short-term fluctuations and long-term offsets of these state quantities, a digital description of the equipment's health status is formed.

[0075] In one possible embodiment, the aforementioned flood warning platform can weight and combine environmental feature vectors and state feature vectors to obtain a feature fusion vector. Specifically, the weights are dynamically allocated based on the correlation coefficient between the environment and the equipment, and a comprehensive feature vector is generated through a fusion algorithm to reflect the overall flood risk trend and to distinguish between normal fluctuations caused by environmental changes and sudden changes caused by equipment abnormalities.

[0076] The aforementioned key characteristic parameters may include, but are not limited to, data on water level change rate, immersion duration, tilt angle change trend, and humidity change amplitude. Specifically, water level change rate reflects the rate of water accumulation; immersion duration indicates the length of time water accumulates; tilt angle change trend reflects the structural stability of the equipment; and humidity change amplitude indicates the degree of abrupt change in environmental humidity. These parameters serve as core inputs for risk assessment, used to determine the development trend and severity of water immersion.

[0077] The aforementioned historical key feature parameter thresholds are statistically derived from historical monitoring data and used to define the boundary between normal and abnormal states. These thresholds can be dynamically adjusted based on factors such as season, geography, and equipment type to maintain the model's adaptability. For example, during the rainy season in southern China (such as the plum rain season), if the aforementioned flood warning platform detects a continuous high humidity environment and still uses the humidity threshold of the dry season (such as 75%), false alarms will frequently be triggered. In this case, the platform automatically increases the humidity threshold to 85%~90% based on historical climate data, making the recognition algorithm more tolerant of natural humidity fluctuations. Conversely, in winter or the dry season, the aforementioned flood warning platform automatically lowers the threshold to enhance sensitivity to abnormal humidity. Alternatively, for coastal or underground areas where baseline humidity and groundwater levels are consistently high, the platform will automatically increase the humidity and water level thresholds based on the geographical identifier of the deployment location. In dry northern regions, the aforementioned flood warning platform lowers the thresholds for judging water level and humidity fluctuations to ensure that abnormal water accumulation events can be detected in a timely manner.

[0078] In this embodiment, the difference between the current key feature parameters and the corresponding historical thresholds can be calculated, and a water immersion risk index can be generated based on the calculation results. This index reflects the current risk level of the target area and serves as an indicator for subsequent anomaly identification and early warning strategy matching.

[0079] Optionally, before comparing the key feature parameters with historical key feature parameter thresholds to generate a flood risk index, the steps may include determining historical monitoring data and corresponding flood risk change trend data within the target area; adjusting the upper and lower limits of the thresholds for the corresponding key feature parameters based on the historical monitoring data and corresponding flood risk change trend data to obtain the target threshold; and updating and replacing the historical key feature parameter thresholds corresponding to the current key feature parameters based on the target threshold.

[0080] In this embodiment of the invention, the aforementioned historical monitoring data may be environmental and equipment operation records stored long-term in the aforementioned flood warning platform, including information such as water level, flood depth, humidity, voltage, and tilt angle at different time periods. These data can be filtered and visualized statistically analyzed by day, week, or month to assess the fluctuation range of characteristic parameters under normal operating conditions.

[0081] The aforementioned data on the changing trends of flood risk can be time-series change information generated based on historical risk events, used to reflect the fluctuation pattern of risk level over time. By analyzing the frequency, duration and risk level distribution of past warnings, risk change patterns under different climatic or geographical conditions are extracted to form trend curves.

[0082] In this embodiment, the aforementioned water immersion early warning platform can automatically adjust the threshold values ​​of each characteristic parameter based on historical monitoring data and risk change trends. Specifically, when a long-term high humidity or a slow upward trend in water level is detected, the corresponding upper limit threshold is automatically increased; conversely, during dry seasons or when the equipment is stable for a long period of time, the lower limit threshold is decreased to enhance the sensitivity of anomaly detection.

[0083] The aforementioned target threshold can be a new upper and lower limit range obtained after adjustment, used to replace the old static threshold. Specifically, the calculated target threshold can be written into the system threshold configuration library to replace the original historical key feature parameter threshold. It is calculated by comprehensively considering the historical fluctuation range, trend slope and abnormal frequency. It can be personalized according to the characteristics of the monitoring area to make the risk model more in line with the actual operating environment.

[0084] By employing the above methods and steps, the risk identification model can adapt to changes in different seasons and environmental conditions during long-term operation, maintaining its sensitivity and stability, reducing false alarm and false negative rates, and improving the reliability and environmental adaptability of the overall early warning effect.

[0085] Optionally, the step of identifying anomalies in water immersion risk data using a preset anomaly algorithm and obtaining anomaly identification results further includes: constructing a corresponding multi-dimensional feature matrix based on the water immersion risk data; calculating the deviation of the multi-dimensional feature matrix to obtain the difference value between each feature and normal data; determining the corresponding anomaly type when the difference value exceeds the preset difference threshold range of the corresponding feature, and classifying the anomaly level according to the size of the difference value; and generating anomaly identification results based on the anomaly type and anomaly level.

[0086] In this embodiment of the invention, after receiving real-time water immersion risk data, a corresponding feature matrix structure can be established according to different feature dimensions, such as water level, humidity, voltage, and tilt angle. For example, a multi-dimensional feature matrix can be constructed based on the time dimension and the sensing feature dimension. At the three sampling times t1, t2, and t3, the water level data of monitoring node N1 are 12cm, 18cm, and 25cm, the humidity is 68%, 75%, and 83%, respectively, the voltage is 220V, 218V, and 217V, and the tilt angle is 0.3°, 0.6°, and 1.2°.

[0087] Arrange these data in rows by sampling time and in columns by feature type such as water level, humidity, voltage, and tilt angle to generate the following structure:

[0088] This table is a multidimensional feature matrix. The aforementioned water immersion early warning platform can perform subsequent deviation calculations based on this matrix. For example, it can compare the feature values ​​at the current time (t3) with the historical average or normal operation model, such as the differences in average water level of 10cm, humidity of 60%, voltage of 220V, and tilt angle of 0°, to determine whether the water level is rising rapidly, whether the humidity is rising continuously, or whether the equipment posture has deviated abnormally.

[0089] In this embodiment, a difference analysis can also be performed between the current monitoring data and the reference model data. Specifically, the aforementioned flood warning platform calculates the magnitude and direction of change of each feature compared to the normal operating state to quantify the degree of deviation. The larger the deviation value, the more significant the difference between the current monitoring data and the standard state, and the higher the risk level.

[0090] The aforementioned difference values, or deviation values, can be used to measure the degree of deviation between each characteristic parameter and the normal baseline. Difference values ​​are typically expressed as standardized proportions or numerical ranges, and are used to determine the severity and scope of the anomaly.

[0091] The aforementioned preset difference threshold range can be an acceptable range of differences set during the modeling phase, used to distinguish between normal fluctuations and abnormal states. Different features have independent thresholds; for example, the allowable range for water level differences is narrower, while the allowable range for tilt angle differences is relatively larger. The aforementioned flood warning platform determines whether an anomaly exists based on whether the difference value exceeds the corresponding threshold range.

[0092] The above abnormal types may include, but are not limited to, water level abnormality (rapid rise of accumulated water); humidity abnormality (sudden increase in ambient humidity); tilt abnormality (device attitude deviation); voltage abnormality (unstable power supply), etc.

[0093] The above abnormal levels can be divided into levels such as prompt, warning, and alarm according to the magnitude of the difference value, which are used to reflect the severity of the abnormality. For example, the above waterlogging warning platform is classified according to the difference amplitude range. The prompt level only records the event, the warning level triggers a notification, and the alarm level activates an audible and visual alarm or linkage control. And after the abnormal type determination and level division are completed, a structured abnormal recognition result is generated, including but not limited to the abnormal type, level, involved characteristic parameters, timestamp, and device number.

[0094] Optionally, in the step of determining the corresponding abnormal type when the difference value exceeds the preset difference threshold range of the corresponding feature and dividing the abnormal level according to the magnitude of the difference value, it further includes tracing the difference value that exceeds the preset difference threshold range of the corresponding feature to determine the corresponding abnormal type; based on the abnormal type, extracting the abnormal change feature data associated with it from the waterlogging risk data; according to the difference value amplitude range corresponding to the abnormal change feature data, and dividing the abnormal level into a prompt level, a warning level, and an alarm level.

[0095] In the embodiment of the present invention, after detecting that the difference value of a certain characteristic parameter exceeds the threshold, the source and formation reason of the abnormal data can be analyzed reversely. For example, when the humidity change value exceeds the set threshold, the above waterlogging warning platform will trace its corresponding monitoring node, water level change record, and communication status, and judge whether the abnormality is caused by environmental change, water level rise, or sensor failure, so as to locate the cause of the abnormality.

[0096] The above abnormal types can be the abnormal attribution categories determined according to the tracing result, which are used to distinguish different types of abnormal situations, including but not limited to water level abnormality (such as a sharp rise in accumulated water); humidity abnormality (such as continuously high humidity); tilt abnormality (such as device attitude deviation or structural deformation); voltage abnormality (such as power supply fluctuation or short-term power failure).

[0097] The above abnormal change feature data can refer to the key parameter sequence related to the abnormal type, which reflects the change law before and after the occurrence of the abnormality. For example, for water level abnormality, the feature data includes the water level rise rate and duration; for tilt abnormality, the feature data includes the angle deviation curve and attitude recovery time; for voltage abnormality, the feature data is the voltage drop rate and recovery fluctuation amplitude.

[0098] The aforementioned range of variation values ​​can refer to the intervals set to distinguish the severity of anomalies based on the magnitude of the deviation. For example, when the rate of water level rise exceeds the threshold by 20% to 40%, it is judged as a minor anomaly; exceeding 40% to 70% is a moderate anomaly; and exceeding 70% is a severe anomaly. This range of variation values ​​is dynamically updated by the aforementioned flood warning platform based on historical monitoring patterns to match the risk tolerance of different environments and equipment characteristics.

[0099] In one possible embodiment, when the difference value exceeds a preset threshold range, the aforementioned water immersion early warning platform performs source tracing analysis to determine the root cause of the anomaly, such as sensor failure, equipment tilting, or a sudden rise in ambient water level. Based on the source tracing results, it determines the anomaly type and extracts corresponding anomaly change characteristic data from the risk data, such as the rate of water level change, the magnitude of humidity increase, voltage fluctuation range, or the duration of tilt angle, to describe the specific manifestation of the anomaly process. The platform then classifies the anomaly based on the range of differences in these characteristic data: a small deviation is a warning level, a moderate deviation is a alert level, and a significantly excessive deviation is an alarm level. The final anomaly identification result includes the anomaly type, level, and triggering characteristics, which drive the execution of subsequent early warning strategies.

[0100] Optionally, the step of executing the corresponding water immersion early warning scheme based on the anomaly identification results and in conjunction with the preset anomaly handling strategy further includes matching the preset anomaly handling strategy according to the anomaly type and anomaly level corresponding to the anomaly identification results; generating and executing the corresponding early warning command based on the matched anomaly handling strategy, the early warning command being used to control the alarm device, remote push terminal or video monitoring module in the target area to execute the early warning response; and recording the abnormal event information and updating or optimizing the anomaly handling strategy after completing the early warning response.

[0101] In this embodiment of the invention, after an anomaly is detected, the anomaly type and anomaly level contained in the anomaly identification result are first read and matched with the preset anomaly handling strategies in the database. For example, when "abnormal water level, alarm level" is detected, the aforementioned water flood warning platform matches the corresponding strategy item "activate audible and visual alarm + video transmission + remote push"; if it is "abnormal humidity, warning level", the matching strategy is "send notification + record event log". The preset strategies are configured by the management terminal and can be flexibly adjusted according to different scenarios and risk levels.

[0102] After matching is complete, the aforementioned flood warning platform automatically generates a set of control commands, including but not limited to control commands transmitted in real time to on-site alarm devices such as audible and visual alarms, buzzers, remote push terminals such as the management center, mobile apps, and video monitoring modules such as cameras. The platform can also send commands to the execution devices via an IoT communication interface, enabling synchronized triggering of on-site alarms, remote alerts, and image linkage. During execution, the platform monitors the command feedback status in real time to ensure successful execution; if the command is not responded to or the device is offline, a retry or backup channel is triggered.

[0103] After the early warning response is completed, the aforementioned flood warning platform automatically records abnormal event information, including trigger time, abnormality type, level, executing device, and processing result.

[0104] The aforementioned flood warning platform updates or optimizes its anomaly handling strategies based on historical event statistics. For example, when a certain type of warning frequently triggers false alarms, the platform can automatically adjust the trigger threshold or delay strategy. When a certain type of alarm is handled effectively, the strategy weight is automatically increased, enabling strategy self-learning and continuous optimization.

[0105] When executing the aforementioned warning, users accessing the flood warning platform also need to ensure data and operational security through a built-in multi-factor authentication mechanism. Specifically, this embodiment adopts a two-factor authentication mechanism of account password + TOTP dynamic verification code. During the user login phase, the user first enters an account and password combination that conforms to a strong password policy (the password must contain uppercase letters, lowercase letters, numbers, and special symbols). After passing the initial verification, the aforementioned flood warning platform requires the user to enter a time-based one-time password (TOTP). The TOTP is generated by a key shared by the user terminal and the server, and the key is stored in a security chip in Base32 encoding format.

[0106] The aforementioned flood warning platform generates a time counter by dividing the current Unix timestamp by a fixed step size (30 seconds). Using this counter and a key as input, it calculates a hash value using the HMAC-SHA1 algorithm. Finally, it extracts a 4-byte portion and takes the modulo to generate a 6-bit dynamic verification code. This verification code is only valid for the current 30 seconds; even if an attacker steals an old verification code, it cannot be reused, thus ensuring session security.

[0107] In addition, the aforementioned flood warning platform supports a limit on the number of incorrect login attempts (after 5 incorrect attempts, users must wait 3 minutes for a reset) and requires users to agree to a privacy policy upon first login. The hierarchical user permission design ensures that only users with specific identities can access protected resources or perform high-privilege operations such as initializing the flood warning platform.

[0108] like Figure 2As shown, this embodiment of the invention also provides a flood warning device 200, which includes: The first determining module 201 is used to determine the environmental data in the target area and the status data of the corresponding monitored target devices; The second determining module 202 is used to determine water immersion risk data based on the environmental data and the status data; The first identification module 203 is used to identify anomalies in the water immersion risk data using a preset anomaly algorithm and obtain anomaly identification results. The early warning module 204 is used to execute the corresponding water immersion early warning scheme based on the anomaly identification results and in combination with the preset anomaly handling strategy.

[0109] Optionally, the first determining module 201 mentioned above includes: The first determining submodule is used to collect environmental data in the target area in real time through a preset environmental sensor to obtain water level data, water immersion depth data and humidity data; The second determining submodule is used to detect the status data of the target device through the device detection module to obtain the corresponding operating voltage, communication connection status and tilt angle data. The third determination submodule is used to collect video stream data of the target device and verify the authenticity of the water level data, water immersion depth data, humidity data, operating voltage, communication connection status and tilt angle data to determine the environmental data in the target area and the status data of the corresponding monitored target device.

[0110] Optionally, the second determining module 202 mentioned above includes: The fourth determining submodule is used to extract features from the environmental data and state data to obtain corresponding environmental feature vectors and state feature vectors. The fifth determining submodule is used to fuse the environmental feature vector and the state feature vector to obtain a feature fusion vector; The sixth determination submodule is used to determine key feature parameters within the target area based on the feature fusion vector. The key feature parameters include water level change rate, immersion time, tilt angle change trend, and humidity change amplitude data. The seventh determination submodule is used to compare the key feature parameters with historical key feature parameter thresholds to generate a water immersion risk index. The eighth determination submodule is used to determine the water immersion risk data within the target area based on the water immersion risk index.

[0111] Optionally, the above-mentioned device further includes: The third determining module is used to determine the historical monitoring data and the corresponding water immersion risk change trend data within the target area; The adjustment module is used to adjust the upper and lower limits of the threshold of the corresponding key feature parameters based on the historical monitoring data and the corresponding water immersion risk change trend data to obtain the target threshold. The update module is used to update and replace the historical key feature parameter thresholds corresponding to the current key feature parameters based on the target threshold.

[0112] Optionally, the first identification module 203 mentioned above includes: The first identification submodule is used to construct a corresponding multidimensional feature matrix based on the water immersion risk data; The second identification submodule is used to perform deviation calculation on the multidimensional feature matrix to obtain the difference value between each feature and normal data. The third identification submodule is used to determine the corresponding anomaly type and classify the anomaly level according to the size of the difference value when the difference value exceeds the preset difference threshold range of the corresponding feature. The fourth identification submodule is used to generate anomaly identification results based on the anomaly type and anomaly level.

[0113] Optionally, the aforementioned third identification submodule includes: The first identification unit is used to trace the source of difference values ​​that exceed the preset difference threshold range of the corresponding feature and determine the corresponding anomaly type; The second identification unit is used to extract anomalous change feature data associated with the anomaly type from the water immersion risk data. The third identification unit is used to classify the anomaly level into a prompt level, a warning level, and an alarm level based on the range of difference values ​​corresponding to the abnormal change feature data.

[0114] Optionally, the aforementioned early warning module 204 includes: The first early warning submodule is used to match a preset exception handling strategy based on the exception type and exception level corresponding to the exception identification result. The second early warning submodule is used to generate and execute corresponding early warning instructions based on the matched anomaly handling strategy. The early warning instructions are used to control the alarm device, remote push terminal or video surveillance module of the target area to execute the early warning response. The third early warning submodule is used to record abnormal event information and update or optimize the abnormal handling strategy after completing the early warning response.

[0115] like Figure 3As shown, this embodiment of the invention also provides an electronic device 300, including a processor, which can execute any of the above-described water immersion warning methods.

[0116] Specifically, it includes a processor 301 and a memory 302, as well as a computer program stored in the memory 302 and capable of running on the processor 301 to execute a flood warning method, wherein: The processor 301 executes the calculator program for the flood warning method stored in the memory 302, performing the following steps: Determine the environmental data in the target area and the status data of the corresponding monitored target devices; Based on the environmental and status data, water immersion risk data is determined; Anomaly identification results are obtained by using a preset anomaly algorithm to identify anomalies in the water immersion risk data. Based on the anomaly identification results and in conjunction with the preset anomaly handling strategy, the corresponding water immersion early warning plan is executed.

[0117] Optionally, the processor 301 executes the environmental data, including water level data, water immersion depth data, and humidity data; the status data includes operating voltage, communication connection status, and tilt angle data; and the determination of the environmental data in the target area and the status data of the corresponding monitored target equipment includes: By using preset environmental sensors, environmental data in the target area is collected in real time to obtain water level data, water immersion depth data, and humidity data; The device detection module detects the status data of the target device to obtain the corresponding operating voltage, communication connection status and tilt angle data. By collecting video stream data from the target device and verifying the authenticity of the water level data, immersion depth data, humidity data, operating voltage, communication connection status, and tilt angle data, the environmental data in the target area and the status data of the corresponding monitored target device are determined.

[0118] Optionally, the processor 301 executes the determination of water immersion risk data based on the environmental data and status data, including: Feature extraction is performed on the environmental data and state data to obtain corresponding environmental feature vectors and state feature vectors; The environmental feature vector and the state feature vector are fused together to obtain the feature fusion vector. Based on the feature fusion vector, key feature parameters within the target area are determined. These key feature parameters include water level change rate, immersion duration, tilt angle change trend, and humidity change amplitude data. The key feature parameters are compared with historical key feature parameter thresholds to generate a water immersion risk index; Based on the water immersion risk index, water immersion risk data within the target area is determined.

[0119] Optionally, before processor 301 performs the step of comparing the key feature parameters with historical key feature parameter thresholds to generate a water immersion risk index, the method further includes: Determine the historical monitoring data and corresponding trend data of water immersion risk within the target area; Based on the historical monitoring data and the corresponding water immersion risk change trend data, the upper and lower limits of the thresholds of the corresponding key feature parameters are adjusted to obtain the target threshold. Based on the target threshold, the historical key feature parameter thresholds corresponding to the current key feature parameters are updated and replaced.

[0120] Optionally, the processor 301 executes the preset anomaly algorithm to perform anomaly identification on the water immersion risk data, and obtains anomaly identification results, including: Based on the water immersion risk data, a corresponding multidimensional feature matrix is ​​constructed; The deviation of the multidimensional feature matrix is ​​calculated to obtain the difference value between each feature and the normal data; When the difference value exceeds the preset difference threshold range of the corresponding feature, the corresponding anomaly type is determined, and the anomaly level is classified according to the size of the difference value. An anomaly identification result is generated based on the anomaly type and anomaly level.

[0121] Optionally, the processor 301 executes the step of determining the corresponding anomaly type and classifying the anomaly level according to the magnitude of the difference value when the difference value exceeds the preset difference threshold range of the corresponding feature, including: For differences exceeding a preset difference threshold range for the corresponding feature, trace the source to determine the corresponding anomaly type; Based on the anomaly type, extract the associated abnormal change feature data from the water immersion risk data; Based on the range of difference values ​​corresponding to the abnormal change characteristic data, the abnormality level is divided into a prompt level, a warning level, and an alarm level.

[0122] Optionally, the processor 301 executes the corresponding water immersion early warning scheme based on the anomaly identification result and in conjunction with a preset anomaly handling strategy, including: Based on the anomaly type and anomaly level corresponding to the anomaly identification results, a preset anomaly handling strategy is matched; Based on the matching anomaly handling strategy, a corresponding early warning instruction is generated and executed. The early warning instruction is used to control the alarm device, remote push terminal or video surveillance module of the target area to execute the early warning response. After completing the early warning response, record the abnormal event information and update or optimize the abnormal handling strategy.

[0123] This invention also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the water immersion warning method or the application-side water immersion warning method provided in this invention, and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0124] Those skilled in the art will understand that implementing all or part of the processes in the above embodiments can be done by a computer program instructing related hardware, and can be stored in a computer-readable storage medium. When executed, the program can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0125] The above description discloses only preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for early warning of flooding, characterized in that, include: Determine the environmental data in the target area and the status data of the corresponding monitored target devices; Based on the environmental and status data, water immersion risk data is determined; Anomaly identification results are obtained by using a preset anomaly algorithm to identify anomalies in the water immersion risk data. Based on the anomaly identification results and in conjunction with the preset anomaly handling strategy, the corresponding water immersion early warning plan is executed.

2. The flood warning method as described in claim 1, characterized in that, The environmental data includes water level data, immersion depth data, and humidity data; the status data includes operating voltage, communication connection status, and tilt angle data; and the environmental data for determining the target area and the status data of the corresponding monitored target equipment include: By using preset environmental sensors, environmental data in the target area is collected in real time to obtain water level data, water immersion depth data, and humidity data; The device detection module detects the status data of the target device to obtain the corresponding operating voltage, communication connection status and tilt angle data. By collecting video stream data from the target device and verifying the authenticity of the water level data, immersion depth data, humidity data, operating voltage, communication connection status, and tilt angle data, the environmental data in the target area and the status data of the corresponding monitored target device are determined.

3. The flood warning method as described in claim 1, characterized in that, The determination of water immersion risk data based on the environmental and status data includes: Feature extraction is performed on the environmental data and state data to obtain corresponding environmental feature vectors and state feature vectors; The environmental feature vector and the state feature vector are fused together to obtain the feature fusion vector. Based on the feature fusion vector, key feature parameters within the target area are determined. These key feature parameters include water level change rate, immersion duration, tilt angle change trend, and humidity change amplitude data. The key feature parameters are compared with historical key feature parameter thresholds to generate a water immersion risk index; Based on the water immersion risk index, water immersion risk data within the target area is determined.

4. The flood warning method as described in claim 3, characterized in that, Before comparing the key feature parameters with historical key feature parameter thresholds to generate a water immersion risk index, the method further includes: Determine the historical monitoring data and corresponding trend data of water immersion risk within the target area; Based on the historical monitoring data and the corresponding water immersion risk change trend data, the upper and lower limits of the thresholds of the corresponding key feature parameters are adjusted to obtain the target threshold. Based on the target threshold, the historical key feature parameter thresholds corresponding to the current key feature parameters are updated and replaced.

5. The flood warning method as described in claim 1, characterized in that, The step of using a preset anomaly algorithm to identify anomalies in the water immersion risk data and obtaining anomaly identification results includes: Based on the water immersion risk data, a corresponding multidimensional feature matrix is ​​constructed; The deviation of the multidimensional feature matrix is ​​calculated to obtain the difference value between each feature and the normal data; When the difference value exceeds the preset difference threshold range of the corresponding feature, the corresponding anomaly type is determined, and the anomaly level is classified according to the size of the difference value. An anomaly identification result is generated based on the anomaly type and anomaly level.

6. The flood warning method as described in claim 5, characterized in that, When the difference value exceeds a preset difference threshold range for the corresponding feature, the corresponding anomaly type is determined, and the anomaly level is classified according to the magnitude of the difference value, including: For differences exceeding a preset difference threshold range for the corresponding feature, trace the source to determine the corresponding anomaly type; Based on the anomaly type, extract the associated abnormal change feature data from the water immersion risk data; Based on the range of difference values ​​corresponding to the abnormal change characteristic data, the abnormality level is divided into a prompt level, a warning level, and an alarm level.

7. The flood warning method as described in claim 1, characterized in that, The step of executing a corresponding water immersion early warning scheme based on the anomaly identification results and in conjunction with a preset anomaly handling strategy includes: Based on the anomaly type and anomaly level corresponding to the anomaly identification results, a preset anomaly handling strategy is matched; Based on the matching anomaly handling strategy, a corresponding early warning instruction is generated and executed. The early warning instruction is used to control the alarm device, remote push terminal or video surveillance module of the target area to execute the early warning response. After completing the early warning response, record the abnormal event information and update or optimize the abnormal handling strategy.

8. A water immersion early warning device, characterized in that, include: The first determining module is used to determine the environmental data in the target area and the status data of the corresponding monitored target devices. The second determining module is used to determine water immersion risk data based on the environmental data and the status data; The first identification module is used to identify anomalies in the water immersion risk data using a preset anomaly algorithm, and obtain anomaly identification results. The early warning module is used to execute the corresponding water immersion early warning scheme based on the anomaly identification results and in combination with the preset anomaly handling strategy.

9. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps of the flood warning method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the flood warning method as described in any one of claims 1 to 7.

Images