A system of smart buoys for navigation monitoring and perimeter alerting, designed to improve navigational safety in the Arctic.

Description

2. Critical issues during interruptions and the inability to guarantee data integrity once communication is restored. From a risk assessment perspective, existing technology often retransmits all raw data to a remote server, which heavily consumes satellite bandwidth and generates high response times, making any assessment impossible during communication interruptions. Regarding environmental adaptability, existing buoys generally use a fixed sensor configuration and static calibration parameters; during inter-latitude or inter-seasonal deployments, sensor drifts and measurement deviations become significant, making multi-zone use impossible without hardware replacement or manual calibration. Content of the invention 10 The objective of the present invention is as follows: to find a solution to the aforementioned problems,to provide a smart buoy system for navigation monitoring and peripheral alerting intended for navigation safety in the Arctic. The technical diagram adopted by the present invention is as follows: a system of intelligent navigation monitoring and peripheral alert buoys intended for navigational safety in the Arctic, comprising: a multi-source sensor data collection module, an intermittent communication data processing module, a peripheral computing and autonomous risk assessment module, a global environment adaptation module, a dynamic parameter recalibration module, and a multi-mode communication and power supply guarantee module; the data output of the multi-source sensor data collection module is connected to the input of the global environment adaptation module, and the output of the global environment adaptation module is connected to the input of the dynamic parameter recalibration module.the calibrated data output of the dynamic parameter recalibration module is connected to the data input of the peripheral computing and autonomous risk determination module25, and the control feedback output of the dynamic parameter recalibration module is sent back to the multi-source sensor data collection module; The output of the peripheral computing and autonomous risk determination module is connected to the local memory and data classification input of the intermittent communication data processing module; 30 BE2026 / 7284 3 The bidirectional receive / transmit port of the intermittent communication data processing module is connected to the communication link port of the multimode communication and power assurance module, while its remote update receive port is connected to the remote parameter input of the dynamic parameter recalibration module. 5 In a preferred implementation, the multi-source sensor data collection module includes: a modular sensor interface,An environmental sensing array, a navigation situation sensing array, and a position and motion sensing array. The modular sensor interface consists of uniform electrical standards and physical connection standards, supporting hot-plugging of temperature sensors, salinity sensors, meteorological sensors, AIS receivers, radar modules, and GPS / Beidou positioning modules. The environmental sensing array includes water temperature probes, conductivity probes, barometers, and anemometers, responsible for collecting raw data on temperature, salinity, atmospheric pressure, and wind speed in the relevant maritime area. The navigation situation sensing array includes an AIS receiver and a phased-array radar. decoding the messages of latitude, longitude, speed and heading of surrounding ships,while the front-endradar provides the range and azimuth echo data of the targets. The set of position and movement sensors includes a dual GPS / Beidou20 positioning module and an inertial measurement unit, the positioning module delivering the latitude and longitude of the buoy every second, and the inertial unit providing the triaxial accelerations. 、as well as triaxial attitude angles. The raw data collected by these three sets of sensors is centralized via the modular sensor interface and transmitted to the global environment adaptation module for environmental compensation and calibration. 25 In a preferred implementation, the intermittent communication data processing module internally includes a local data memory, a recovery state machine automatic, a data priority classification table and a synchronization verification engine. The local data memory consists of a high-reliability, solid-state storage assembly BE2026 / 7284 4,recording in full all navigation and environmental data collected by the sensors during communication link interruptions, with time-stamping and a cyclical recovery strategy during writing. The storage capacity is designed for continuous and consistent collection over 180 days. The data priority classification table divides the data to be retransmitted into three levels: emergency alert, critical status, and normal logging. The emergency alert level corresponds to threat alert messages produced by the peripheral computing module and autonomous risk assessment; the critical status level corresponds to summary information on the buoy's position, speed, and hydro-meteorological conditions; the normal logging level corresponds to complete sensor sampling sequences and system operating logs. In a preferred implementation, the automatic recovery state machine maintains a record of the data transfer progress for each transmission link,including the sequence numbers of the data packets confirmed as received and the number of the next packet to be transmitted. After the communication link is re-established, it automatically initiates a data index comparison request with the ground center, and after receiving the list of missing packets, these are transmitted in descending order of priority. The synchronization verification engine waits for acknowledgment from the ground center after sending each batch of data packets, and for segments that failed or have verification errors, performs a retransmission up to a maximum of three times; packets exceeding this limit are relegated to the end of the regular log queue for the next synchronization cycle. In a preferred implementation, the edge computing and autonomous risk determination module internally includes an edge computing engine, comprising: a data preprocessing chain,a multi-source fusion analysis core and an alert message generator. The hardware unit of the peripheral computing engine consists of an embedded processor and a dedicated neural network accelerator, directly executing real-time alert analysis tasks on the terminal, without retransmitting raw sensor data to a remote server. The data preprocessing chain aligns timestamps and eliminates outliers from calibrated data coming from the parameter dynamics recalibration module; the multi-source fusion analysis core receives AIS target trajectories, radar points, and hydro-meteorological data streams to perform spatial association and feature vector extraction. The message generator The alert system immediately constructs alert frames after triggering the fusion analysis and transmits them to the data processing module via intermittent communication for high priority processing. In a preferred implementation,The peripheral calculation and autonomous risk assessment module uses the peripheral calculation engine as hardware support and receives the stream of calibrated multi-source data from the dynamic parameter recalibration module, including: AIS data from surrounding vessels (latitude, ground speed, heading), the distances and azimuth angles of targets tracked by the radar module, as well as the real-time position P_buoy and the movement vector M_buoy of the buoy provided by the GPS and the inertial unit of measurement. The module first performs a spatio-temporal alignment, unifying the AIST_AIS data update period and the radar scan period T_radar on a time base of 1 second, and converts the geographic coordinates into a system of 15 local north-east coordinates centered on the buoy according to the WGS-84 ellipsoidal model. Then, it calculates the nearest passing distance D_CPA and the nearest passing time T_CPA between the buoy and the target vessels, the distance D_CPA being determined by geometric projection of the relative motion,using the difference of velocity vectors ΔV and the difference of position vectors ΔP to solve the distance closure factors and the plane intersection relationships. To account for the particularities of the Arctic environment, the module introduces the ice density factor ρ_ice and the ice zone navigation correction coefficient k_ice, extending the standard collision risk determination threshold of 2 nautical miles to = + ×. The module also calculates the buoy drift risk index R_drift, normalized from the triaxial standard deviation of the acceleration σ_acce and the position variance σ_pos collected by the inertial unit, combined with the maximum drift radius r_max allowed in the maritime zone. For the detection of environmental changes, the module uses the temporal gradient of pressure p, temperature T_air and wind speed W_speed; if the pressure variation dp / dt exceeds the critical threshold of 300 Pa / h,A BE2026 / 7284 6 environmental risk marker is triggered. All these quantified results are then integrated into a weighted risk matrix fusion model to produce a global risk level R_total, with local classification of the alert level and generation of corresponding messages. The calculation model for the adaptive collision risk threshold in ice zones is as follows: = ×1 + ×5 where: actual is the safe passage distance threshold actually used in current ice conditions, expressed in nautical miles; base is the standard safe passage distance threshold in open waters, set at 2.0 nautical miles; is the dimensionless correction coefficient for navigation in ice zones, between 0.5 and 1.5, determined according to the ship's ice class and icebreaker escort status according to the reference table; is the dimensionless ice floe density factor, between 0 and 1,obtained from satellite data on the ice or ice cover measured by local visual sensors.15 The model for calculating the drift risk index based on the uncertainty of the movement is as follows: = √ 2+ 2 where: is the dimensionless drift risk index, between 0 and 1, the closer its value is to 1, the higher the drift risk;20 is the standard deviation of the triaxial acceleration over the sliding time window, expressed in meters per square second, reflecting the intensity of the instantaneous movements of the buoy; is the standard deviation of the GPS position over the same time window, expressed in meters, reflecting the dispersion of the cumulative drift of the buoy; is the maximum permissible drift radius of the buoy's anchoring system, expressed in 25 meters, predefined according to the length of the anchor chain and the water depth. The calculation model for the overall risk index by weighted fusion of a multi-factor risk matrix is ​​as follows: BE2026 / 7284 7 = 1× + 2× + 3× where: is the overall risk index, dimensionless; is the collision risk sub-index,obtained by mapping using a fuzzy membership function from D_CPA and T_CPA; is the drift risk index; 5 is the sub-index of risk related to environmental changes, calculated from the temporal rates of variation of environmental parameters such as the pressure gradient and the wind speed gradient, normalized to thresholds; 1, 2, 3 are weighting coefficients, respecting the normalization constraint 1+2+3=1, and dynamically adjusted according to the type of maritime zone identified by the 10 global environment adaptation module. In ice zones, the values ​​of 2 and 3 are higher than those of tropical shipping lanes, reflecting the relative importance given to drift risk and environmental changes according to the maritime zone. In a preferred implementation, the global environment adaptation module internally includes a unit for identifying the type of marine environment,a controller for matching model parameters and an adaptation layer for sensor interfaces. The marine environment type identification unit uses the buoy's current latitude, temperature, and salinity as input characteristics and classifies the environment into predefined types such as high-latitude ice zone, temperate zone, and busy tropical shipping lane, by matching partable. The matching result is updated every 20 hours based on the buoy's movement. The model parameter matching controller stores an environmental compensation parameter model for each marine zone type, including temperature compensation coefficients, salinity calibration offsets, and atmospheric pressure correction factors, and automatically loads the compensation model for the new zone when the identified zone type changes, by sending the switching instruction to the dynamic parameter recalibration module. In a preferred implementation,The sensor interface adaptation layer of the global environment adaptation module is defined by uniform electrical standards and physical interface standards, connecting all sensors of the multi-source data collection module. The interface layer executes a sensor driver management program responsible for identifying the present sensor model and loading the corresponding communication chip. When an environmental mode change occurs or a sensor is replaced with a different model for maintenance, the interface adaptation layer notifies the dynamic parameter recalibration module of the change event, which loads the five new calibration parameters corresponding to the new configuration. In a preferred implementation, the dynamic parameter recalibration module internally includes a dynamic environmental parameter compensation algorithm unit.A sensor calibration instruction generator and a remote parameter update receiver. The dynamic compensation unit receives the 10-parameter compensation model for the current marine area provided by the global environment adaptation module and performs online compensation of the raw data from the multi-source data collection module, including temperature drift correction, cross-compensation for salinity and conductivity, and pressure-depth conversion. The compensated data is transmitted to the autonomous peripheral risk assessment module for risk evaluation. The sensor calibration instruction generator produces recalibration instruction frames according to a predefined cycle, based on the correction calculated by the algorithm and the detected zero offset of the sensor, and sends them to the corresponding sensors via the interface adaptation layer. In a preferred implementation,The multimode communication and power supply module includes internally a satellite communication unit, a VHF communication unit, a data transmission arbiter, and a power supply unit. The satellite communication unit, consisting of LEO / Iridium or Tiangong satellite transmit / receive modules, ensures the long-distance transmission of data and the reception of instructions from the ground center in open waters. The VHF communication unit, consisting of a VHF transceiver and an antenna, provides a fast, short-range data exchange channel near coastlines or in areas with high ship density. The data transmission arbiter monitors in real time the signal strength and error rate of the satellite and VHF links, and makes switching decisions between link quality and power consumption. prioritizing the sending of high-priority data on the link whose design power exceeds a predefined threshold, and in case of failure of both links,sends a local memory trigger signal to the data processing module in intermittent communication. The power unit consists of photovoltaic solar panels, lithium batteries, and a power management circuit; the solar panels charge the batteries via an MPPT controller, and the management circuit converts the battery voltage to power the five other modules. When the battery level is below a predefined threshold, the transmission arbiter automatically limits the transmission of non-urgent data to extend battery life. In summary, thanks to the adoption of the technical solutions described above, the beneficial effects of the present invention are as follows: 1. In the present invention, thanks to the collaboration between local data memory and the resumption algorithm after interruption, when extreme environments such as arctic magnetic storms or intense tropical storms cause a complete interruption of satellite communications,The buoy fully records all navigation and environmental data via the local storage module. Data writing uses timestamp marking and a cyclic recovery strategy, ensuring that no monitoring information is lost during the interruption. Once the communication link is re-established, the automatic recovery state machine automatically initiates a data index comparison request to the ground center, synchronizing the missing data packets according to priority order: emergency alert, important state, normal logging. After sending each batch of packets, the synchronization verification engine waits for acknowledgment from the ground center and then retransmits the failed segments. This mechanism guarantees that, even under intermittent communication conditions, the ground center eventually obtains the 1. The complete data sequence, and that critical alert information is transmitted with priority without being affected by the accumulation of ordinary data. 2. In the present invention,The introduction of the edge computing module allows the system to perform real-time analysis of raw sensor data directly on the terminal. The edge computing engine executes the autonomous risk matrix determination algorithm, autonomously evaluating the risk of collision, abnormal drift, or threats of environmental change, instead of retransmitting the raw data to a remote server for processing. Once the evaluation is complete, only alert messages and summaries of critical situations are sent back, while the numerous ordinary sampling sequences and system logs do not occupy the real-time communication bandwidth. This architecture fundamentally reduces data transfer. unnecessary, reduces dependence on both the remote server's computing resources and satellite bandwidth, and reduces the satellite round-trip alert response time to 5 times the local latency.ensuring that the evaluation capacity is not affected during satellite interruptions. 3. In the present invention, the global environment adaptation module and the dynamic parameter recalibration module, thanks to the modular sensor interface at the hardware level and the dynamic environmental parameter compensation algorithm at the software level, give the system applicability to all areas, from high-latitude ice zones to busy tropical shipping lanes. The modular sensor interface, complying with uniform electrical and physical standards, supports hot-plugging of components such as temperature sensors, salinity sensors, and meteorological sensors, allowing the configuration of sensor combinations according to the characteristics of the target area before buoy deployment. The dynamic environmental parameter compensation algorithm identifies the type of area based on latitude, of salinity and temperature, loads the corresponding compensation parameter model,andappliesthermaldriftcorrection,salinity-conductivitycrosscompensationandpressure-depthconversiontotherawsensordata,whilesendingcalibrationinstructionframes20forreal-timecalibration.Thisarchitectureallowsthesamehardwareandbuoyalgorithmtomaintaindatacollectionaccuracyandriskassessmentreliabilityindifferentmarineenvironments,withouttheneedforspecificdevelopmentorrecalibrationforeacharea. Descriptionofthedrawings25 Figure1isaglobalfunctionaldiagramofthesystemofthepresentinvention; Figure2isafunctionaldiagramoftheperipheralcalculationandautonomousriskdeterminationmoduleofthepresentinvention.SpecificEmbodimentIn ordertofurtherclarifytheobjective,technicaldiagramandadvantagesofthepresent30 BE2026 / 7284 11 invention,This is explained in detail below with reference to the figures and examples of implementation. It should be understood that the specific examples described herein are intended only to illustrate the present invention and do not limit it. Referring to Figures 1 and 2, a smart buoy system for navigation monitoring and perimeter alerting intended for navigational safety in the Arctic comprises: 5 a multi-source sensor data collection module, an intermittent communication data processing module, a perimeter computing and autonomous risk assessment module, a global environment adaptation module, a dynamic parameter recalibration module, and a multi-mode communication and power assurance module. 10 The data output of the multi-source sensor data collection module is connected to the input of the global environment adaptation module; the output of the global environment adaptation module is connected to the input of the dynamic parameter recalibration module.and the calibrated data output of the latter is connected to the data input of the peripheral calculation and autonomous risk determination module.15 The control feedback output of the dynamic parameter recalibration module is returned to the multi-source sensor data collection module. The output of the peripheral computing and autonomous risk determination module is connected to the local memory and data classification input of the intermittent communication data processing module. The bidirectional receive / transmit port of the intermittent communication data processing module is connected to the communication link port of the multimode communication and power assurance module, while its remote update receive port is connected to the remote parameter input of the dynamic parameter recalibration module. The multi-source data collection module includes: a modular sensor interface, an environmental perception sensor set,a set of navigation situation perception sensors and a set of position and movement sensors. The modular sensor interface consists of uniform electrical and connection standards, supporting hot-plugging of temperature sensors, 30 BE2026 / 7284 12 salinity sensors, weather sensors, AIS receivers, radar modules, and GPS / Beidou positioning modules. The environmental sensing sensor suite includes water temperature probes, conductivity probes, barometers, and anemometers, responsible for collecting raw data on temperature, salinity, atmospheric pressure, and wind speed in the maritime area. The navigation situation sensing sensor suite includes an AIS receiver and a phased-array radar front-end, the AIS receiver decoding the latitude, longitude, speed, and heading messages of surrounding vessels.and the front-endradar providing target distance and azimuth echo data. The position and motion sensor suite includes a dual GPS / Beidou positioning module and an inertial measurement unit. The positioning module delivers the buoy's latitude and longitude every second, and the inertial unit provides triaxial accelerations and triaxial attitude angles. The raw data collected by these three suites are centralized via the modular sensor interface and transmitted to the global environment adaptation module for compensation and environmental calibration. The intermittent communication data processing module internally comprises a local data memory, an automatic restart state machine, a data priority classification table, and a synchronization verification engine. The local memory consists of a high-reliability solid-state storage array.recording in full all navigation and environmental data collected by the sensors during communication interruptions. Data writing uses a timestamp marking system and a cyclic recovery strategy, with a storage capacity designed for continuous and consistent collection over 180 days. The priority classification table divides the data to be retransmitted into three levels: emergency alert, critical status, and normal logging. The emergency alert level corresponds to threat alert messages produced by the peripheral computing module, the critical status level corresponds to summary information on the buoy's position, speed, and hydro-meteorological conditions, and the normal logging level corresponds to complete sampling sequences and system logs. The automatic recovery state machine maintains a record of the progress of the BE2026 / 7284 13 data transfer for each link,including the sequence numbers of the confirmed data packets and the number of the next packet to be transmitted. After the communication link is re-established, it automatically initiates a data index comparison request with the ground center, and after receiving the list of missing packets, the packets are retransmitted in descending priority order. The synchronization verification engine waits for acknowledgment from the ground center after sending each batch of packets and performs the retransmission of the failed segments, with a maximum of three attempts. Packets exceeding this limit are relegated to the end of the regular log queue for the next synchronization cycle. The edge computing and autonomous risk assessment module includes a 10 edge computing engine, comprising: a data preprocessing chain, a multi-source fusion analysis kernel, and an alert message generator. The engine hardware unit consists of an embedded processor and a dedicated neural network accelerator.directly executing real-time alert analysis tasks on the terminal at sea, without retransmitting raw data to a remote server. The preprocessing chain aligns the 15 timestamps and eliminates outliers from the calibrated data coming from the dynamic parameter recalibration module, while the multi-source fusion analysis core receives AIS target trajectories, radar points, and hydro-meteorological data streams to perform spatial association and feature vector extraction. The alert message generator immediately constructs the alert frames after the fusion analysis is triggered and transmits them to the data processing module via intermittent communication for high-priority processing. The peripheral calculation and autonomous risk determination module uses the peripheral calculation engine as hardware support and receives the calibrated multi-source data stream from the dynamic parameter recalibration module, including: AIS data from the 25 surrounding ships (latitude, ground speed, heading),The distances and azimuth angles of the targets tracked by the radar module, as well as the real-time position P_buoy and the motion vector M_buoy of the buoy provided by the GPS and the inertial unit of measurement. The module first performs a spatio-temporal alignment, unifying the data update period AIST_AIS and the radar scan period T_radar on a time base of 1 second, and converts the geographic coordinates into a local northeast coordinate system centered on the buoy according to the WGS-84 ellipsoidal model. Then, it calculates the nearest passing distance D_CPA and the nearest passing time T_CPA between the buoy and the target vessels, D_CPA being determined by geometric projection of the relative motion from the difference between the velocity vectors ΔV and the relative position vectors ΔP to solve the closure tensors of the 5 distances and the plane intersection relations. To take into account the particularities of the Arctic environment, the module introduces the ice density factor ρ_ice and the ice zone navigation correction coefficient k_ice,extending the standard collision risk determination threshold from 2 nautical miles to: = + ×. The module also calculates the drift risk index R_drift, normalized from the triaxial standard deviation of 10, the acceleration σ_accet and the position variance σ_pos measured by the inertial unit, combined with the maximum drift radius allowed in the maritime zone. For the detection of environmental changes, the module uses the time gradient of pressure p, temperature T_aire and wind speed W_speed; if the pressure variation dp / dt exceeds the critical threshold of 300 Pa / h, an environmental risk marker is triggered. All these quantified results are then integrated into a weighted risk matrix fusion model to produce a global risk level R_total, with local classification of alert level and generation of corresponding messages. The model for calculating the adaptive collision risk threshold in icy conditions is as follows: =×1+×20 where: is the safe passage distance threshold actually used in current icy conditions,expressed in nautical miles; is the standard safe passage distance threshold in open waters, set at 2.0 nautical miles; is the dimensionless ice zone correction coefficient, between 0.5 and 1.5, determined according to the ship's ice class and icebreaker escort status according to the reference table; is the dimensionless ice density factor, between 0 and 1, obtained at BE2026 / 7284 15 from satellite data on ice or ice cover measured by local visual sensors. The model for calculating the drift risk index based on the uncertainty of the movement is as follows: = √2 + 2 5 where: is the dimensionless drift risk index, between 0 and 1, the closer its value is to 1, the higher the drift risk; is the standard deviation of the triaxial acceleration over the sliding time window, expressed in meters per square second, reflecting the intensity of the instantaneous movements of the buoy; is the standard deviation of the GPS position over the same time window, expressed in meters,10 reflecting the cumulative drift dispersion of the buoy; is the maximum permissible drift radius of the buoy's anchoring system, expressed in meters, predefined according to the length of the anchor chain and the water depth. The calculation model for the overall risk index by weighted fusion of a multi-factor risk matrix is ​​as follows: 15 = 1× + 2× + 3× where: Rtotal is the dimensionless overall risk index; Rcollision is the collision risk sub-index, obtained by mapping using a fuzzy membership function from D_CPA and T_CPA; is the drift risk index; 20 is the risk subindex linked to environmental changes, calculated from the temporal variation rates of environmental parameters such as the pressure gradient and the wind speed gradient, normalized according to the thresholds; 1、2、3 are weighting coefficients, respecting the normalization constraint 1+2+3=1, dynamically adjusted according to the type of maritime zone identified by module 25 of adaptation to the global environment. In ice zones,The values ​​of 2 and 3 are higher than those of tropical shipping lanes, reflecting the relative importance given to drift risk and environmental changes depending on the maritime zone. BE2026 / 7284 16 The global environment adaptation module internally comprises a marine environment type identification unit, a model parameter matching controller, and a sensor interface adaptation layer. The identification unit uses the buoy's current latitude, temperature, and salinity as input characteristics and classifies the environment into predefined types such as high-latitude ice zone, temperate zone 5, and busy tropical shipping lane, using a shared matching system. The matching results are updated hourly based on the buoy's movement. The matching controller stores, for each marine zone type, a model of environmental compensation parameters, including temperature compensation coefficients.The salinity calibration offsets and atmospheric pressure correction factors, and 10 automatically load the corresponding compensation model when the identified zone type changes, by sending the switching instruction to the dynamic parameter recalibration module. The sensor interface adaptation layer of the adaptation module is defined by uniform electrical standards and physical interface standards, connecting all the sensors 15 of the multi-source data collection module. The interface layer executes a sensor driver management program responsible for identifying the present sensor model and loading the corresponding communication stack. When an environmental mode change occurs or a sensor is replaced by a different model for maintenance, the adaptation layer notifies the dynamic parameter recalibration module of the event.which loads the new calibration parameters corresponding to the new configuration. The dynamic parameter recalibration module internally comprises a dynamic environmental parameter compensation algorithm unit, a sensor calibration instruction generator, and a remote parameter update receiver. The compensation unit receives the current marine zone compensation parameter model provided by the adaptation module and performs online compensation of the raw data from the multi-source data collection module, including temperature drift correction, salinity-conductivity cross-compensation, and pressure-depth conversion. The compensated data is transmitted to the peripheral calculation module for risk assessment. The instruction generator produces, according to the calculated correction and the zero offset of the 30 BE2026 / 7284 17 sensor,Recalibration instruction frames are generated according to a predefined cycle and sent to the sensors via the adaptation layer. The remote parameter update receiver is connected to the update instruction channel of the data processing module in intermittent communication. During the communication interruption period, it receives the algorithm parameter update packets 5 sent by the ground center, including the new compensation curves for the maritime zones and the reference values ​​for sensor recalibration. Once received, the parameters of the dynamic compensation module are placed online and the update status is sent back to the ground center for confirmation. The module also contains a calibration log recorder that saves, on an event-driven basis, the timestamp of each recalibration,the correction amplitude and the response state of the sensors. This log is part of the normal logging level data and is transmitted to the ground center during the next synchronization cycle of the data processing module for remote diagnosis and long-term drift trend analysis. The multimode communication and power supply module internally comprises 15 a satellite communication unit, a VHF communication unit, a data transmission arbiter, and a power supply unit. The satellite communication unit consists of LEI / Iridium or Tiangong satellite transmit / receive modules, ensuring the long-distance transmission of data and the reception of instructions from the ground center in open waters. The VHF communication unit, composed of a transceiver and an antenna, 20 provides a fast data exchange channel over short distances near coastlines or in areas with high ship density. The transmission arbiter monitors in real time the signal strength and error rates of the satellite and VHF links.It makes the switching decisions between link quality and energy consumption, prioritizing the transmission of high-priority data on the link whose design power exceeds a predefined threshold, and, if both links are interrupted, sends a local memory trigger signal to the intermittent communication data processing module. The power unit includes photovoltaic solar panels, lithium batteries, and a power management circuit; the solar panels charge the batteries via an MPPT controller, and the management circuit converts the battery voltage to power the five other modules. When the battery level is below a predefined threshold, the transmission arbiter automatically limits the transmission of non-urgent data to extend autonomy. Thus, in the present invention, thanks to the cooperation between the local data memory and the resume after interruption algorithm,When extreme environments such as arctic magnetic storms or intense tropical storms cause a complete interruption of satellite communications, the buoy fully records all navigation and environmental data in the local storage module, with time-stamping and a cyclic recovery strategy, ensuring that no monitoring information is lost during the interruption. Once the communication link is re-established, the automatic recovery state machine initiates a data index comparison request to the ground center, synchronizing the missing packets according to priority order: emergency alert, critical status, normal logging. After sending each batch of packets, the synchronization verification engine awaits acknowledgment from the ground center and retransmits the failed segments, ensuring that the ground center obtains the complete sequence. data,and that critical alert information be transmitted as a priority without being affected by the accumulation of ordinary data. The introduction of the peripheral computing module allows the system to perform real-time analysis of raw sensor data directly on the terminal. The computing engine executes the autonomous risk matrix determination algorithm, autonomously evaluating the risk of collision, abnormal drift, or threats of environmental changes, instead of retransmitting the raw data to a remote server. Once the evaluation is complete, only alert messages and summaries of critical situations are transmitted, while the numerous routine sampling sequences and system logs do not occupy the real-time communication bandwidth. This architecture fundamentally reduces the transfer of unnecessary data and decreases dependence on computing resources. from the remote server and satellite bandwidth, the response time for alerts is reduced from the satellite round-trip delay to local latency.ensuring that the evaluation capacity is not affected during satellite outages. The global environment adaptation module and the dynamic parameter recalibration module, thanks to the modular sensor interface at the hardware level and the dynamic environmental parameter compensation algorithm at the software level, give the system applicability across all areas, from high-latitude ice zones to busy tropical shipping lanes. The modular interface supports hot-plugging of temperature, salinity, and meteorological sensors, allowing for co-,