Installation buoy state monitoring method for marine environment monitoring based on Internet of Things

Abstract

The invention discloses an installation buoy state monitoring method for marine environment monitoring based on the Internet of Things, and relates to the field of the Internet of Things, and the method comprises the steps: laying a buoy group according to a task demand, and building a communication link; self-inspection is started, parameters are configured according to needs, and the buoy state and environment data are collected in combination with a periodic mode and a trigger mode; after being preprocessed and transmitted to a monitoring platform through multi-channel redundancy, the data is subjected to space-time fusion and standardization processing; the intelligent analysis module is used for evaluating the health condition and environmental adaptability of the buoy so as to recognize abnormity, early warning is given out according to risk classification, and linkage response is started; the system continuously accumulates data, and continuously optimizes an identification model and an alarm threshold by means of a gradient descent method to realize closed-loop optimization and dynamic parameter updating. The system has the advantages that by means of the Internet of Things sensing network and the adaptive multi-channel redundancy transmission and intelligent analysis technology, comprehensive and accurate buoy state monitoring in the complex marine environment is achieved, communication is stable, and operation and maintenance are efficient.

Description

Technical Field

[0001] This invention relates to the field of the Internet of Things (IoT), and in particular to a method for monitoring the status of installed buoys for marine environmental monitoring based on the IoT. Background Technology

[0002] Currently, the reliable operation of marine environmental monitoring buoys is crucial for obtaining continuous and accurate ocean data. The development of their status monitoring methods stems from the inherent limitations of traditional manual inspection methods, such as high cost, high risk, and poor real-time performance, which cannot meet the urgent need for refined management of large-scale buoy arrays.

[0003] Current marine buoy status monitoring methods on the market mostly employ a single point-to-point communication network architecture, lacking edge computing units and a three-tiered sensing network design. In environments with high salt spray, strong winds and waves, and interference from ships, connection interruptions are easily caused by protocol incompatibility and signal blockage, with single-network outages in remote sea areas averaging over 180 hours per year. Data acquisition often adopts a fixed-cycle mode without a trigger-based dynamic adjustment mechanism, and single-channel transmission lacks redundancy backup. Due to satellite bandwidth limitations or electromagnetic interference, data packet loss rates are high, and some systems have up to 30% invalid data due to lack of preprocessing. Analysis and early warning rely on single parameter threshold judgments, failing to correlate marine environment with buoy operational status, making it difficult to distinguish between natural displacement and equipment failure, resulting in high false alarm rates and delayed responses. Positioning accuracy significantly decreases under severe weather conditions. The lack of a closed-loop iterative optimization mechanism and fixed parameter configurations make it impossible to adapt to differences in water depth and tidal currents in different sea areas. Historical data is not effectively reused, leading to insufficient long-term operational stability, high maintenance costs, and an inability to meet the precise and long-term requirements of marine environmental monitoring. Summary of the Invention

[0004] To improve existing methods, this paper provides an IoT-based method for monitoring the status of installed buoys for marine environmental monitoring. This method relies on a three-level IoT sensing network, integrates multi-sensor fusion acquisition, adaptive multi-channel redundant transmission and intelligent analysis technology, and is coupled with a hierarchical early warning and closed-loop iterative optimization mechanism to achieve comprehensive and accurate buoy status monitoring, stable communication and efficient operation and maintenance in complex marine environments.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] The IoT-based method for monitoring the status of installed buoys in marine environmental monitoring includes:

[0007] Based on the marine environmental characteristics of the target sea area and the monitoring mission requirements, the coordinates and distribution density of the buoys are obtained. The IoT sensing nodes integrated on each buoy are connected to the monitoring platform through the IoT communication module, and a three-level IoT sensing network of buoy end-communication link-monitoring platform is constructed.

[0008] By sending initialization commands to the IoT sensing nodes at the buoy end, the self-test programs of each sensor, communication module and positioning module are started, and core parameters, including data acquisition frequency, data transmission protocol and positioning update frequency, are configured based on the monitoring task requirements.

[0009] The buoy's operational status data is collected in real time through a status monitoring sensor group, and the marine environment data of the buoy's location is collected synchronously through an environmental parameter sensor group. A data collection mode combining periodic and trigger-based data collection is adopted.

[0010] After being preprocessed, the data cached locally at the buoy end is transmitted to the monitoring platform through the Internet of Things communication module, and the main communication channel is adaptively selected based on the sea area communication signal coverage, and multi-channel redundant transmission is performed.

[0011] Based on the received buoy operational status data and marine environmental data, time synchronization and spatial correlation are performed through a spatiotemporal fusion algorithm, and standardized transformation is performed according to a preset data model to obtain a standardized dataset;

[0012] Based on standardized datasets, the intelligent analysis module of the monitoring platform performs multi-dimensional analysis of buoy status, including operational health analysis, environmental adaptability analysis, and abnormal status identification.

[0013] Based on the anomaly identification results, the warning levels are divided according to the severity of the anomalies. The corresponding warning information is pushed to the designated terminal through the Internet of Things communication link, and the linkage response mechanism is activated.

[0014] The monitoring platform accumulates buoy operational status data, marine environmental data, and early warning response results, builds a historical database, uses a gradient descent algorithm to iteratively optimize the anomaly identification model parameters and early warning thresholds, and synchronizes the optimized parameters to IoT sensing nodes of the same type of buoy.

[0015] Preferably, the step of obtaining the coordinates and distribution density of buoy deployment based on the marine environmental characteristics of the target sea area and the monitoring task requirements, and establishing a connection between the IoT sensing nodes integrated on each buoy and the monitoring platform through the IoT communication module to construct a three-level IoT sensing network of buoy-communication link-monitoring platform specifically includes:

[0016] Based on the water depth distribution data of the target sea area, and combined with the route planning map, the number of passing vessels and the route distribution are counted daily to obtain the deployment coordinates and distribution density of each buoy.

[0017] The sensing node includes a status monitoring sensor group, an environmental parameter sensor group, an Internet of Things communication module, and a GNSS positioning module;

[0018] The buoy sends an initial connection request to the monitoring platform via the buoy communication module. After receiving the request, the platform assigns a unique communication link identifier and connects the output of each sensing node to the buoy edge computing unit via an RS485 bus to aggregate sensor data and build a three-level IoT sensing network of buoy-communication link-monitoring platform.

[0019] Preferably, the step of sending an initialization command to the IoT sensing node at the buoy end to initiate the self-test program of each sensor, communication module, and positioning module, and configuring core parameters based on the monitoring task requirements, including data acquisition frequency, data transmission protocol, and positioning update frequency, specifically includes:

[0020] The monitoring platform sends an initialization command containing the module startup sequence to the buoy end, starts the self-test program of each sensor, communication module and positioning module, and verifies the hardware operating status and the validity of data acquisition.

[0021] Zero-point calibration and range calibration are performed on the condition monitoring sensor group and the environmental parameter sensor group to eliminate inherent equipment errors;

[0022] Configure core parameters based on monitoring task requirements, including data acquisition frequency, data transmission protocol, and location update frequency;

[0023] The unique identifier of the buoy is bound to the user account of the monitoring platform, and simulated status data is sent to the buoy to perform node and platform identity authentication and communication link testing.

[0024] Preferably, the method of collecting buoy operational status data in real time through a status monitoring sensor group and synchronously collecting marine environmental data of the buoy's location through an environmental parameter sensor group, employing a combination of periodic and trigger-based acquisition, specifically includes:

[0025] The operational status data includes the communication module signal strength, the stress value of the buoy body structure, the buoy attitude angle, and the GNSS positioning coordinates;

[0026] The environmental data includes sea surface temperature, seawater salinity, significant wave height, wave period, and ocean current speed and direction.

[0027] A combination of periodic and trigger-based data collection is adopted, and the collection frequency is increased when the monitored data exceeds the preset value.

[0028] The collected data is cached in the buoy's local storage module, with a cache capacity of 72 hours of continuously collected data.

[0029] Preferably, the data locally cached at the buoy end is preprocessed and then transmitted to the monitoring platform via the IoT communication module. The main communication channel is adaptively selected based on the sea area's communication signal coverage, and multi-channel redundant transmission is specifically implemented including:

[0030] The data cached locally at the buoy end is preprocessed by data format standardization and invalid data filtering, and then transmitted to the cloud monitoring platform through the Internet of Things multi-communication module. The communication channels include satellite communication channel, 4G / 5G mobile communication channel, and LoRa low-power wide area network channel.

[0031] Real-time monitoring of signal strength for each channel; when the 4G / 5G channel signal is ≥-90dBm, it is set as the main channel; when the signal is <-90dBm, it switches to the satellite communication channel; the LoRa channel is used as a near-shore auxiliary backup.

[0032] During transmission, the LZ77 data compression algorithm is used to reduce the amount of data transmitted, and the data is encrypted using the AES-128 encryption algorithm.

[0033] Preferably, the step of synchronizing and spatially correlating the received buoy operational status data and marine environmental data using a spatiotemporal fusion algorithm, and then standardizing the data according to a preset data model to obtain a standardized dataset specifically includes:

[0034] Extract the timestamps of data from each sensor, align the data collected by different sensors using a time synchronization algorithm, and bind environmental parameters to their corresponding spatial locations based on the buoy's GNSS positioning coordinates.

[0035] For normally distributed data, the 3σ criterion is used to remove outliers; for data with large fluctuations, the Grubbs criterion is used to identify outliers, and linear interpolation is used to complete missing data.

[0036] Based on a pre-defined model and a unified data format, the data is converted into a Parquet columnar storage structure, and a label is added to each parameter.

[0037] The integrated and processed data is used to generate a standardized dataset, with the buoy ID and collection timestamp as a combined index.

[0038] Preferably, the step of performing multi-dimensional analysis of the buoy status based on a standardized dataset through the intelligent analysis module of the monitoring platform, including operational health analysis, environmental adaptability analysis, and abnormal status identification, specifically includes:

[0039] The operational health analysis assesses the working status of the communication module based on the frequency and duration of daily signal strength below -105dBm and the interruption duration, and judges the structural integrity of the buoy body based on the trend of structural stress value changes.

[0040] The environmental adaptability analysis assesses the impact of environmental factors on the buoy by correlating marine environmental data with buoy operational status data.

[0041] The abnormal state identification process extracts buoy state feature parameters and combines them with a machine learning model for feature matching and anomaly determination to obtain the abnormal state type and the location and time of the anomaly.

[0042] Preferably, the step of classifying warning levels according to the severity of the anomaly based on the anomaly identification results, pushing warning information of the corresponding level to designated terminals through the IoT communication link, and activating the linkage response mechanism specifically includes:

[0043] Based on the anomaly identification results, the warning levels are divided according to the severity of the anomalies, including mild warning, moderate warning, and severe warning;

[0044] Abnormal data is pushed to the mobile phones of maintenance personnel via IoT links, displaying real-time curves of abnormal parameters and historical comparison charts, and triggering audible and visual alarms.

[0045] During a mild warning, the buoy hibernation trigger time is reduced from 30 minutes to 20 minutes, and the data collection frequency is adjusted from 5 minutes / time to 8 minutes / time. During a moderate warning, a scheduling notification containing a maintenance list is sent, and the optimal route is marked. During a severe warning, the buoy's real-time GNSS trajectory is pushed to the maintenance vessel, which then arrives at the site for maintenance or recovery.

[0046] Preferably, the monitoring platform accumulates buoy operational status data, marine environmental data, and early warning response results, constructs a historical database, iteratively optimizes the anomaly identification model parameters and early warning thresholds using a gradient descent algorithm, and synchronizes the optimized parameters to IoT sensing nodes of the same type of buoy. Specifically, this includes:

[0047] The monitoring platform continuously accumulates data on buoy operational status, marine environmental data, and early warning response results to build a historical database;

[0048] Based on historical data, the gradient descent algorithm is used to iteratively optimize the parameters of the anomaly identification model and fine-tune the warning threshold.

[0049] Based on the differences in environmental characteristics of different sea areas and the dynamic changes in monitoring tasks, the monitoring scheme is adaptively optimized by remotely adjusting the data acquisition frequency and communication protocol parameters.

[0050] The optimized parameters are synchronized to the IoT sensing nodes of the same type of buoy, forming a closed-loop iterative mechanism of data collection, analysis, optimization and application.

[0051] Compared with the prior art, the advantages of the present invention are:

[0052] A fully intelligent and highly reliable buoy status monitoring system has been constructed: A three-tiered IoT sensing network—from the buoy terminal to the communication link and the monitoring platform—combined with multi-sensor fusion and self-calibration mechanisms, ensures comprehensive monitoring coverage and accurate data acquisition. A combined periodic and trigger-based acquisition mode, coupled with local 72-hour caching, LZ77 compression, and AES-128 encryption technology, balances data continuity and transmission security. Adaptive selection of 4G / 5G, satellite, and LoRa multi-channel redundant transmission based on marine signal characteristics guarantees communication stability in complex marine environments. Spatiotemporal fusion algorithms and multi-dimensional intelligent analysis accurately assess buoy operational health and environmental adaptability, and identify anomalies. A tiered early warning and linkage response mechanism enables precise and efficient anomaly handling, while closed-loop iterative optimization driven by gradient descent algorithms, combined with historical databases, continuously adjusts parameters to adapt to different marine environments and monitoring needs, significantly improving the intelligence level, operational efficiency, and long-term operational reliability of buoy monitoring, providing stable, accurate, and sustainable technical support for marine environmental monitoring. Attached Figure Description

[0053] Figure 1 This is a schematic diagram of the method proposed in this invention;

[0054] Figure 2 This is a schematic diagram of the IoT sensing network construction proposed in this invention;

[0055] Figure 3 This is a schematic diagram of the initialization of the sensing node proposed in this invention;

[0056] Figure 4 This is a schematic diagram of the data acquisition and local caching proposed in this invention;

[0057] Figure 5 This is a schematic diagram of the multi-channel redundant data transmission proposed in this invention;

[0058] Figure 6 This is a schematic diagram illustrating the data fusion and standardization proposed in this invention;

[0059] Figure 7 This is a schematic diagram of the intelligent analysis and anomaly identification proposed in this invention;

[0060] Figure 8 This is a schematic diagram of the hierarchical early warning system proposed in this invention;

[0061] Figure 9 This is a schematic diagram of the iterative optimization and parameter adjustment proposed in this invention. Detailed Implementation

[0062] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0063] See Figure 1 As shown, the method for monitoring the status of installed buoys for marine environmental monitoring based on the Internet of Things includes:

[0064] Step 1: Based on the marine environmental characteristics of the target sea area and the monitoring task requirements, obtain the coordinates and distribution density of the buoy deployment, and establish a connection between the IoT sensing nodes integrated on each buoy and the monitoring platform through the IoT communication module, thus constructing a three-level IoT sensing network of buoy end - communication link - monitoring platform.

[0065] Step 2: By sending an initialization command to the IoT sensing node at the buoy end, the self-test program of each sensor, communication module and positioning module is started, and the core parameters are configured based on the monitoring task requirements, including data acquisition frequency, data transmission protocol and positioning update frequency.

[0066] Step 3: Collect buoy operational status data in real time through the status monitoring sensor group, and synchronously collect marine environmental data of the buoy's location through the environmental parameter sensor group, using a combination of periodic and trigger-based acquisition modes;

[0067] Step 4: After preprocessing, the data locally cached at the buoy is transmitted to the monitoring platform through the IoT communication module, and the main communication channel is adaptively selected based on the sea area communication signal coverage, and multi-channel redundant transmission is performed;

[0068] Step 5: Based on the received buoy operational status data and marine environmental data, time synchronization and spatial correlation are performed using a spatiotemporal fusion algorithm, and standardized transformation is performed according to a preset data model to obtain a standardized dataset;

[0069] Step Six: Based on the standardized dataset, the buoy status is analyzed from multiple dimensions through the intelligent analysis module of the monitoring platform, including operational health analysis, environmental adaptability analysis, and abnormal status identification;

[0070] Step 7: Based on the anomaly identification results, classify the warning levels according to the severity of the anomalies, push the corresponding warning information to the designated terminals through the IoT communication link, and activate the linkage response mechanism;

[0071] Step 8: The monitoring platform accumulates buoy operation status data, marine environment data, and early warning response results, builds a historical database, uses the gradient descent algorithm to iteratively optimize the anomaly identification model parameters and early warning thresholds, and synchronizes the optimized parameters to the IoT sensing nodes of the same type of buoy.

[0072] See Figure 2As shown, based on the marine environmental characteristics of the target sea area and the monitoring mission requirements, the coordinates and distribution density of the buoys are obtained. The IoT communication module connects the IoT sensing nodes integrated on each buoy to the monitoring platform, constructing a three-tiered IoT sensing network: buoy-communication link-monitoring platform. Specifically, this includes:

[0073] Based on the water depth distribution data of the target sea area, and combined with the route planning map, the number of passing vessels and the route distribution are counted daily to obtain the deployment coordinates and distribution density of each buoy.

[0074] The sensing node includes a status monitoring sensor group, an environmental parameter sensor group, an Internet of Things communication module, and a GNSS positioning module;

[0075] The buoy sends an initial connection request to the monitoring platform via the buoy communication module. After receiving the request, the platform assigns a unique communication link identifier and connects the output of each sensing node to the buoy edge computing unit via an RS485 bus to aggregate sensor data and build a three-level IoT sensing network of buoy-communication link-monitoring platform.

[0076] Specifically, during the installation of the condition monitoring sensors, the structural stress sensor is attached to the main structural component 50cm above and below the buoy's waterline, and the attitude sensor is fixed at the buoy's central axis. Among the environmental parameter sensors, the temperature and salinity sensors are installed inside the sensor sleeve 1m underwater, and the wave and current sensors are installed on the outer support of the buoy. The communication module and GNSS module are integrated on the top of the buoy.

[0077] See Figure 3 As shown, by sending an initialization command to the IoT sensing node at the buoy end, the self-test program of each sensor, communication module, and positioning module is initiated. Based on the monitoring task requirements, core parameters are configured, including data acquisition frequency, data transmission protocol, and positioning update frequency. Specifically, these include:

[0078] The monitoring platform sends an initialization command containing the module startup sequence to the buoy end, starts the self-test program of each sensor, communication module and positioning module, and verifies the hardware operating status and the validity of data acquisition.

[0079] Zero-point calibration and range calibration are performed on the condition monitoring sensor group and the environmental parameter sensor group to eliminate inherent equipment errors;

[0080] Configure core parameters based on monitoring task requirements, including data acquisition frequency, data transmission protocol, and location update frequency;

[0081] The unique identifier of the buoy is bound to the user account of the monitoring platform, and simulated status data is sent to the buoy to perform node and platform identity authentication and communication link testing.

[0082] Specifically, the data acquisition frequency is set in layers according to the monitored objects. Under normal conditions, the communication signal strength is collected once every 5 minutes, the structural stress and buoy attitude are collected once every 10 minutes, the seawater temperature and salinity are collected once every 15 minutes, and the waves and ocean currents are collected once every 20 minutes. When any parameter exceeds the preset normal range, the high-frequency acquisition mode is automatically triggered, and the acquisition frequency of the corresponding sensor is increased to once per minute. The communication protocol is adaptively configured according to the signal coverage of the sea area. The 4G / 5G HTTP protocol is enabled by default in the near-shore area, and the satellite communication MQTT protocol is automatically switched in the far-sea area where there is no mobile signal. The LoRa CoAP protocol is enabled in the shallow sea near-distance networking scenario. The protocol switching threshold is set to start switching when the signal strength is <-100dBm.

[0083] The unique identifier of the buoy is encrypted and bound to the user account on the cloud platform. The binding process uses two-way authentication. The buoy uploads the hardware feature code, and the cloud platform issues a binding confirmation command after verification to complete the identity association. Then, the communication link stability test is started, and test data packets are sent continuously for 12 hours. The test indicators include data packet reception success rate, transmission delay, and bit error rate. If data packets are lost or the delay exceeds the standard, the transmission power of the communication module is automatically adjusted and the test is repeated until the link stability requirements are met.

[0084] See Figure 4 As shown, the buoy's operational status data is collected in real time through a status monitoring sensor group, and the marine environment data of the buoy's location is collected synchronously through an environmental parameter sensor group. The data collection mode, which combines periodic and triggered data collection, specifically includes:

[0085] The operational status data includes the communication module signal strength, the stress value of the buoy body structure, the buoy attitude angle, and the GNSS positioning coordinates;

[0086] The environmental data includes sea surface temperature, seawater salinity, significant wave height, wave period, and ocean current speed and direction.

[0087] A combination of periodic and trigger-based data collection is adopted, and the collection frequency is increased when the monitored data exceeds the preset value.

[0088] The collected data is cached in the buoy's local storage module, with a cache capacity of 72 hours of continuously collected data.

[0089] Specifically, a collaborative mechanism of periodic and trigger-based data acquisition is implemented, with preset trigger thresholds: when the communication signal strength is ≤-110dBm for 3 consecutive times, the structural stress is ≥120MPa, the attitude angle fluctuation exceeds 30°, or the seawater temperature changes suddenly by ≥5℃ / hour, or the significant wave height is ≥8m, the high-frequency acquisition mode is automatically activated, and the corresponding sensor acquisition frequency is increased to 1-3 minutes / time, while recording the timestamp of the trigger event and the associated environmental parameters; the trigger state continues until the parameters return to the normal range, and then the high-frequency acquisition is maintained for 30 minutes before resuming the normal frequency;

[0090] First, standardize the format by using a structured format of buoy ID-acquisition timestamp-parameter type-value-unit, with timestamps accurate to milliseconds and parameter units standardized to international standard units. Then, activate a two-layer filtering mechanism: the first layer is validity filtering, which removes invalid data with missing sensor signals for more than three acquisition cycles and overflow data exceeding the sensor's range; the second layer is interference filtering, which identifies abnormal data caused by electromagnetic interference or marine organism attachment by judging the parameter change rate, marks it as pending verification, and temporarily stores it, excluding it from the main transmission and analysis process.

[0091] See Figure 5 As shown, the data locally cached at the buoy end is preprocessed and then transmitted to the monitoring platform via the IoT communication module. The main communication channel is adaptively selected based on the sea area's communication signal coverage, and multi-channel redundant transmission is specifically implemented, including:

[0092] The data cached locally at the buoy end is preprocessed by data format standardization and invalid data filtering, and then transmitted to the cloud monitoring platform through the Internet of Things multi-communication module. The communication channels include satellite communication channel, 4G / 5G mobile communication channel, and LoRa low-power wide area network channel.

[0093] Real-time monitoring of signal strength for each channel; when the 4G / 5G channel signal is ≥-90dBm, it is set as the main channel; when the signal is <-90dBm, it switches to the satellite communication channel; the LoRa channel is used as a near-shore auxiliary backup.

[0094] During transmission, the LZ77 data compression algorithm is used to reduce the amount of data transmitted, and the data is encrypted using the AES-128 encryption algorithm.

[0095] Specifically, channel priorities are set based on the characteristics of the marine communication environment. In near-shore areas, 4G / 5G mobile communication channels are used as the main channels, and LoRa low-power wide area networks are used as backup channels. In offshore areas, BeiDou satellite communication channels are used as the main channels, and 4G / 5G channels are used as backup channels. In extreme marine environments, a dual backup channel redundancy mode of satellite + LoRa is activated.

[0096] The main channel communication status is monitored in real time. When the main channel detects a signal strength < -105dBm, a transmission delay > 15s, or a bit error rate > 0.05% for three consecutive times, the channel switching process is automatically triggered. During the switching process, the data to be transmitted is temporarily stored in the buffer queue, and the exact switching time is ≤ 2s. After the switch, the backup channel is automatically promoted to the main channel. The original main channel continues to monitor the signal status. When the indicators return to normal and stabilize for 3 minutes, it automatically switches back to the original main channel, forming a dynamic cycle.

[0097] The preprocessed data is compressed using an improved LZ77 compression algorithm, and the compression logic is optimized to address the repetitive characteristics of marine environmental parameters. The encryption process employs the AES-128 symmetric encryption algorithm, with the key dynamically generated by combining the buoy's unique ID with a real-time timestamp, and automatically updated every 24 hours. An encryption identifier and key version number are embedded in the transmitted data packets. To address the high latency of satellite channel transmission, data packet fragmentation technology is used, splitting the large dataset into fragments ≤1KB, each with a fragment sequence number and checksum.

[0098] See Figure 6 As shown, based on the received buoy operational status data and marine environmental data, a spatiotemporal fusion algorithm is used for time synchronization and spatial correlation. Then, according to a preset data model, a standardized dataset is obtained, specifically including:

[0099] Extract the timestamps of data from each sensor, align the data collected by different sensors using a time synchronization algorithm, and bind environmental parameters to their corresponding spatial locations based on the buoy's GNSS positioning coordinates.

[0100] For normally distributed data, the 3σ criterion is used to remove outliers; for data with large fluctuations, the Grubbs criterion is used to identify outliers, and linear interpolation is used to complete missing data.

[0101] Based on a pre-defined model and a unified data format, the data is converted into a Parquet columnar storage structure, and a label is added to each parameter.

[0102] The integrated and processed data is used to generate a standardized dataset, with the buoy ID and collection timestamp as a combined index.

[0103] Specifically, to address the issues of asynchronous data acquisition and installation deviations in spatial location among multiple sensors in marine monitoring, a spatiotemporal fusion algorithm is employed for precise calibration. Regarding time synchronization, the timestamp of the GNSS positioning module is used as a benchmark to correct the offset of acquisition times for state and environmental sensors. For example, the acquisition times of structural stress and temperature sensors are aligned to the nearest GNSS time node, eliminating time differences caused by sensor response delays. For spatial calibration, the data acquired by each sensor is mapped to the buoy's center coordinate reference point by combining the buoy's dimensions and sensor installation location parameters, correcting spatial deviations caused by different installation locations and ensuring spatiotemporal consistency of multi-dimensional data from the same buoy at the same time point.

[0104] See Figure 7 As shown, based on a standardized dataset, the intelligent analysis module of the monitoring platform performs multi-dimensional analysis of the buoy's status, including operational health analysis, environmental adaptability analysis, and abnormal status identification. Specifically, this includes:

[0105] The operational health analysis assesses the working status of the communication module based on the frequency and duration of daily signal strength below -105dBm and the interruption duration, and judges the structural integrity of the buoy body based on the trend of structural stress value changes.

[0106] The environmental adaptability analysis assesses the impact of environmental factors on the buoy by correlating marine environmental data with buoy operational status data.

[0107] The abnormal state identification process extracts buoy state feature parameters and combines them with a machine learning model for feature matching and anomaly determination to obtain the abnormal state type and the location and time of the anomaly.

[0108] Specifically, the intelligent analysis module constructs a three-dimensional health assessment model based on a standardized dataset. The health of the communication module focuses on the amplitude of signal strength fluctuations, the frequency of interruptions, and the success rate of protocol switching. The working status is determined by statistically analyzing data from the past 24 hours. The structural health combines the peak stress above and below the waterline, the duration of stress, and the tolerance threshold of the buoy material. If the peak stress exceeds 180MPa and lasts for more than 5 minutes, the structural risk is directly marked. All indicators are correlated with historical data from the same period.

[0109] The formula for scoring the structural health of a buoy is:

[0110]

[0111] in, To score the structural health, This represents the peak structural stress at a point 50cm above and below the waterline over the past 24 hours. The stress tolerance threshold of the buoy material. The duration of the stress peak. This is the threshold for the allowable duration of peak stress.

[0112] By employing feature matching and scene mapping logic, the intrinsic correlation between the marine environment and the buoy's state is explored. For example, the data on significant wave height and ocean current velocity are time-series aligned with the data on structural stress and attitude angle to establish a correspondence between environmental intensity and state response: when the significant wave height is ≥6m, the structural stress is allowed to increase to 150MPa and the attitude angle fluctuation is allowed to be ≤45°. If the values ​​exceed this range, they are judged as abnormal. Based on the environmental characteristics of different sea areas, differentiated correlation thresholds are preset to improve the adaptability of the analysis.

[0113] See Figure 8 As shown, based on the anomaly identification results, warning levels are divided according to the severity of the anomaly. Corresponding warning information is pushed to designated terminals via the IoT communication link, and a linkage response mechanism is activated, specifically including:

[0114] Based on the anomaly identification results, the warning levels are divided according to the severity of the anomalies, including mild warning, moderate warning, and severe warning;

[0115] Abnormal data is pushed to the mobile phones of maintenance personnel via IoT links, displaying real-time curves of abnormal parameters and historical comparison charts, and triggering audible and visual alarms.

[0116] During a mild warning, the buoy hibernation trigger time is reduced from 30 minutes to 20 minutes, and the data collection frequency is adjusted from 5 minutes / time to 8 minutes / time. During a moderate warning, a scheduling notification containing a maintenance list is sent, and the optimal route is marked. During a severe warning, the buoy's real-time GNSS trajectory is pushed to the maintenance vessel, which then arrives at the site for maintenance or recovery.

[0117] Specifically, the cloud-based monitoring platform classifies warning levels based on anomaly identification results, according to a three-dimensional standard of impact range, severity of harm, and recovery difficulty. Mild warnings target anomalies that do not affect core monitoring functions, such as communication signal fluctuations ≤ 5 times per day; moderate warnings target anomalies that affect data quality, such as communication interruptions 6-10 times per day or structural stress of 150-180 MPa; severe warnings target anomalies that threaten buoy safety, such as structural stress ≥ 180 MPa or attitude angle tilt exceeding 45°. Each level corresponds to a structured instruction, including core fields such as execution subject, operation type, parameter threshold, and completion time limit. For example, a mild warning instruction explicitly states the buoy end: adjust power strategy, reduce non-core sensor acquisition frequency to 20 minutes / time, and complete within 10 minutes; a severe warning instruction specifies the maintenance team: on-site inspection, lock buoy coordinates, and depart within 24 hours.

[0118] See Figure 9As shown, the monitoring platform accumulates buoy operational status data, marine environmental data, and early warning response results, constructs a historical database, and uses a gradient descent algorithm to iteratively optimize the anomaly identification model parameters and early warning thresholds. The optimized parameters are then synchronized to IoT sensing nodes of similar buoys. Specifically, this includes:

[0119] The monitoring platform continuously accumulates data on buoy operational status, marine environmental data, and early warning response results to build a historical database;

[0120] Based on historical data, the gradient descent algorithm is used to iteratively optimize the parameters of the anomaly identification model and fine-tune the warning threshold.

[0121] Based on the differences in environmental characteristics of different sea areas and the dynamic changes in monitoring tasks, the monitoring scheme is adaptively optimized by remotely adjusting the data acquisition frequency and communication protocol parameters.

[0122] The optimized parameters are synchronized to the IoT sensing nodes of the same type of buoy, forming a closed-loop iterative mechanism of data collection, analysis, optimization and application.

[0123] Specifically, based on historical archived data covering datasets from different sea areas, seasons, and fault types, the intelligent analysis model is updated regularly. Model training employs a combination of incremental training and full optimization. Incremental training is performed quarterly using newly added data to correct model parameters. Full optimization is performed annually by integrating more than three years of full data, optimizing feature weight allocation, such as increasing the weight of features related to salt spray corrosion in high-salinity sea areas. Model validation uses a two-dimensional standard: offline validation evaluates performance by comparing the recognition accuracy and false positive rate of historical fault data; validation selects 10% of buoys as pilots to test the optimized model, track the timeliness of anomaly recognition and the accuracy of early warning, and optimize the anomaly judgment threshold. Based on the environmental baselines of different sea areas, differentiated threshold standards are formulated to improve model adaptability.

[0124] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0125] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0126] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for monitoring the status of installed buoys for marine environmental monitoring based on the Internet of Things, characterized in that, include: Based on the marine environmental characteristics of the target sea area and the monitoring mission requirements, the coordinates and distribution density of the buoys are obtained. The IoT sensing nodes integrated on each buoy are connected to the monitoring platform through the IoT communication module, and a three-level IoT sensing network of buoy end-communication link-monitoring platform is constructed. By sending initialization commands to the IoT sensing nodes at the buoy end, the self-test programs of each sensor, communication module and positioning module are started, and core parameters, including data acquisition frequency, data transmission protocol and positioning update frequency, are configured based on the monitoring task requirements. The buoy's operational status data is collected in real time through a status monitoring sensor group, and the marine environment data of the buoy's location is collected synchronously through an environmental parameter sensor group. A data collection mode combining periodic and trigger-based data collection is adopted. After being preprocessed, the data cached locally at the buoy end is transmitted to the monitoring platform through the Internet of Things communication module, and the main communication channel is adaptively selected based on the sea area communication signal coverage, and multi-channel redundant transmission is performed. Based on the received buoy operational status data and marine environmental data, time synchronization and spatial correlation are performed through a spatiotemporal fusion algorithm, and standardized transformation is performed according to a preset data model to obtain a standardized dataset; Based on standardized datasets, the intelligent analysis module of the monitoring platform performs multi-dimensional analysis of buoy status, including operational health analysis, environmental adaptability analysis, and abnormal status identification. Based on the anomaly identification results, the warning levels are divided according to the severity of the anomalies. The corresponding warning information is pushed to the designated terminal through the Internet of Things communication link, and the linkage response mechanism is activated. The monitoring platform accumulates buoy operational status data, marine environmental data, and early warning response results, builds a historical database, uses a gradient descent algorithm to iteratively optimize the anomaly identification model parameters and early warning thresholds, and synchronizes the optimized parameters to IoT sensing nodes of the same type of buoy.

2. The IoT-based installation buoy state monitoring method for ocean environment monitoring according to claim 1, characterized by, Based on the marine environmental characteristics of the target sea area and the monitoring mission requirements, the coordinates and distribution density of the deployed buoys are obtained. The IoT communication module connects the integrated IoT sensing nodes on each buoy to the monitoring platform, constructing a three-tiered IoT sensing network: buoy-communication link-monitoring platform. This specifically includes: Based on the water depth distribution data of the target sea area, and combined with the route planning map, the number of passing vessels and the route distribution are counted daily to obtain the deployment coordinates and distribution density of each buoy. The sensing node includes a status monitoring sensor group, an environmental parameter sensor group, an Internet of Things communication module, and a GNSS positioning module; The buoy sends an initial connection request to the monitoring platform via the buoy communication module. After receiving the request, the platform assigns a unique communication link identifier and connects the output of each sensing node to the buoy edge computing unit via an RS485 bus to aggregate sensor data and build a three-level IoT sensing network of buoy-communication link-monitoring platform. 3.The IoT-based installation buoy state monitoring method for ocean environment monitoring according to claim 1, wherein, The process involves sending initialization commands to the IoT sensing nodes at the buoy end to initiate self-test procedures for each sensor, communication module, and positioning module. Based on the monitoring task requirements, core parameters are configured, including data acquisition frequency, data transmission protocol, and positioning update frequency. Specifically, this includes: The monitoring platform sends an initialization command containing the module startup sequence to the buoy end, starts the self-test program of each sensor, communication module and positioning module, and verifies the hardware operating status and the validity of data acquisition. Zero-point calibration and range calibration are performed on the condition monitoring sensor group and the environmental parameter sensor group to eliminate inherent equipment errors; Configure core parameters based on monitoring task requirements, including data acquisition frequency, data transmission protocol, and location update frequency; The unique identifier of the buoy is bound to the user account of the monitoring platform, and simulated status data is sent to the buoy to perform node and platform identity authentication and communication link testing. 4.The IoT-based installation buoy state monitoring method for ocean environment monitoring according to claim 1, wherein, The process of collecting buoy operational status data in real time through a status monitoring sensor group and synchronously collecting marine environmental data of the buoy's location through an environmental parameter sensor group, employing a combined periodic and trigger-based acquisition mode, specifically includes: The operational status data includes the communication module signal strength, the stress value of the buoy body structure, the buoy attitude angle, and the GNSS positioning coordinates; The environmental data includes sea surface temperature, seawater salinity, significant wave height, wave period, and ocean current speed and direction. A combination of periodic and trigger-based data collection is adopted, and the collection frequency is increased when the monitored data exceeds the preset value. The collected data is cached in the buoy's local storage module, with a cache capacity of 72 hours of continuously collected data. 5.The IoT-based installation buoy state monitoring method for ocean environment monitoring according to claim 4, characterized in that, The data locally cached at the buoy end is preprocessed and then transmitted to the monitoring platform via the IoT communication module. The main communication channel is adaptively selected based on the sea area's communication signal coverage, and multi-channel redundant transmission is specifically implemented, including: The data cached locally at the buoy end is preprocessed by data format standardization and invalid data filtering, and then transmitted to the cloud monitoring platform through the Internet of Things multi-communication module. The communication channels include satellite communication channel, 4G / 5G mobile communication channel, and LoRa low-power wide area network channel. Real-time monitoring of signal strength for each channel; when the 4G / 5G channel signal is ≥-90dBm, it is set as the main channel; when the signal is <-90dBm, it switches to the satellite communication channel; the LoRa channel is used as a near-shore auxiliary backup. During transmission, the LZ77 data compression algorithm is used to reduce the amount of data transmitted, and the data is encrypted using the AES-128 encryption algorithm. 6.The IoT-based mounted buoy state monitoring method for ocean environment monitoring according to claim 1, wherein, The process of synchronizing and spatially correlating the received buoy operational status data and marine environmental data using a spatiotemporal fusion algorithm, and then standardizing the data according to a preset data model to obtain a standardized dataset specifically includes: Extract the timestamps of data from each sensor, align the data collected by different sensors using a time synchronization algorithm, and bind environmental parameters to their corresponding spatial locations based on the buoy's GNSS positioning coordinates. For normally distributed data, the 3σ criterion is used to remove outliers; for data with large fluctuations, the Grubbs criterion is used to identify outliers, and linear interpolation is used to complete missing data. Based on a pre-defined model and a unified data format, the data is converted into a Parquet columnar storage structure, and a label is added to each parameter. The integrated and processed data is used to generate a standardized dataset, with the buoy ID and collection timestamp as a combined index. 7.The IoT-based mounted buoy state monitoring method for ocean environment monitoring according to claim 1, wherein, The aforementioned multi-dimensional analysis of buoy status based on standardized datasets, conducted through the intelligent analysis module of the monitoring platform, includes operational health analysis, environmental adaptability analysis, and abnormal status identification. Specifically, this includes: The operational health analysis assesses the communication module's working status based on the frequency of daily signal strength below -105dBm and the duration of interruptions, and judges the structural integrity of the buoy body based on the trend of structural stress value changes. The environmental adaptability analysis assesses the impact of environmental factors on the buoy by correlating marine environmental data with buoy operational status data. The abnormal state identification process extracts buoy state feature parameters and combines them with a machine learning model for feature matching and anomaly determination to obtain the abnormal state type and the location and time of the anomaly. 8.The IoT-based mounted buoy state monitoring method for ocean environment monitoring according to claim 1, wherein, The process of classifying warning levels according to the severity of the anomalies based on anomaly identification results, pushing corresponding warning information to designated terminals via IoT communication links, and activating a coordinated response mechanism specifically includes: Based on the anomaly identification results, the warning levels are divided according to the severity of the anomalies, including mild warning, moderate warning, and severe warning; Abnormal data is pushed to the mobile phones of maintenance personnel via IoT links, displaying real-time curves of abnormal parameters and historical comparison charts, and triggering audible and visual alarms. During a mild warning, the buoy hibernation trigger time is reduced from 30 minutes to 20 minutes, and the data collection frequency is adjusted from 5 minutes / time to 8 minutes / time. During a moderate warning, a scheduling notification containing a maintenance list is sent, and the optimal route is marked. During a severe warning, the buoy's real-time GNSS trajectory is pushed to the maintenance vessel, which then arrives at the site for maintenance or recovery. 9.The IoT-based mounted buoy state monitoring method for ocean environment monitoring according to claim 1, wherein, The monitoring platform accumulates buoy operational status data, marine environmental data, and early warning response results, constructs a historical database, iteratively optimizes anomaly identification model parameters and early warning thresholds using a gradient descent algorithm, and synchronizes the optimized parameters to IoT sensing nodes of similar buoys. Specifically, this includes: The monitoring platform continuously accumulates data on buoy operational status, marine environmental data, and early warning response results to build a historical database; Based on historical data, the gradient descent algorithm is used to iteratively optimize the parameters of the anomaly identification model and fine-tune the warning threshold. Based on the differences in environmental characteristics of different sea areas and the dynamic changes in monitoring tasks, the monitoring scheme is adaptively optimized by remotely adjusting the data acquisition frequency and communication protocol parameters. The optimized parameters are synchronized to the IoT sensing nodes of the same type of buoy, forming a closed-loop iterative mechanism of data collection, analysis, optimization and application.

Images