Large ocean data buoy safety monitoring system and method based on multi-source fusion

Abstract

The application discloses a large ocean data buoy safety monitoring system and method based on multi-source fusion, relates to the technical field of buoy monitoring, and synchronously collects buoy equipment operation parameters, marine environment data and surrounding ship dynamic information through a buoy body sensor group, an environment monitoring unit and a ship identification module; time-frequency domain feature extraction and abnormality detection are carried out on the collected data, a collision risk dynamic evaluation model is constructed in combination with ship trajectory prediction; an equipment health, environment interference and ship threat fusion evaluation system is established, risk grade division is carried out in combination with real-time monitoring data and a historical database; early warning information is generated based on the risk grade, and equipment maintenance instructions and ship collision avoidance alarms are automatically triggered. The application is suitable for the scene of variable speed and steering of a ship, threshold values are dynamically adjusted, and the adaptability and analysis accuracy of the system are improved.

Description

Technical Field

[0001] This invention relates to the field of buoy monitoring technology, specifically to a large-scale marine data buoy safety monitoring system and method based on multi-source fusion. Background Technology

[0002] The marine buoy online monitoring system utilizes advanced Internet of Things (IoT) and wireless sensing technologies to monitor parameters such as water quality, ocean currents, waves, and weather conditions online, providing comprehensive data support for marine ecological research. The observation points use the buoy as a platform, and in addition to being deployed at fixed points on the surface, some instruments can also move vertically via an automatic lifting system, thereby enabling the collection of stratified data.

[0003] In existing technologies, early warning systems for ship collision risks using marine buoys are typically based on fixed distance thresholds, failing to consider real-time threat changes caused by adjustments in ship speed and course. In adverse sea conditions, buoy position drift due to waves can easily trigger warnings. Furthermore, equipment health assessments rely on manual experience to set alarm thresholds, making it difficult to capture early degradation characteristics such as abnormal vibration patterns and gradual changes in power consumption. This also hinders the establishment of nonlinear correlations between equipment performance degradation and multi-source monitoring data, leading to delayed fault warnings. Traditional environmental impact analyses also remain limited to judging single parameter exceedances, weakening adaptability to complex environments.

[0004] Therefore, this invention discloses a large-scale marine data buoy safety monitoring system and method based on multi-source fusion to solve the above problems. Summary of the Invention

[0005] The purpose of this invention is to provide a safety monitoring system and method for large-scale marine data buoys based on multi-source fusion, so as to solve the problems raised in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for safety monitoring of large marine data buoys based on multi-source fusion, the method comprising the following steps: S1: Simultaneously collect buoy equipment operating parameters, marine environmental data, and dynamic information of surrounding vessels through the buoy body sensor group, environmental monitoring unit, and ship identification module; S2: Extract time-frequency domain features and detect anomalies from the collected data, and construct a dynamic assessment model for collision risk by combining ship trajectory prediction; S3: Establish a fusion assessment system for equipment health, environmental interference, and ship threats, and classify risk levels by combining real-time monitoring data and historical databases; S4: Generates early warning information based on risk level, automatically triggering equipment maintenance instructions and ship collision avoidance alarms.

[0007] According to the above scheme, S1 includes the following: S101: Acquire buoy body monitoring data, the buoy body monitoring data including triaxial vibration acceleration a collected by a vibration sensor. x a y a z The system acquires buoy power consumption sequence P(t) via a current sensor, where t represents time; records buoy position drift ΔL via a GPS module; acquires environmental monitoring data, including wave height Hs and wave period TP obtained via wave radar; collects water temperature Tw, salinity S, and water depth D via a temperature, salinity, and depth meter; acquires ship data, including ship position, ship speed vs, and ship heading αs obtained by parsing AIS signals; and extracts ship draft ds through visual recognition. S102: Perform time synchronization and alignment on heterogeneous data; establish a spatial grid mapping model to convert ship positions into polar coordinates (rs, θs) relative to buoys; generate a timestamped fused dataset D={D b D e D v}, where D b For buoy body data, D e For environmental data, D v For ship data.

[0008] This application collects combined monitoring data on vibration, power consumption, and position drift to accurately capture early signs of equipment mechanical failure. It also uses time synchronization to address the phase difference between waves, ship motion, and equipment status. Polar coordinate mapping converts the ship's position into a buoy relative motion model, simplifying the collision prediction calculation dimension. The timestamped fusion dataset provides standardized input for subsequent time-frequency analysis, avoiding conflicts between multiple data source formats.

[0009] According to the above scheme, S2 includes the following: S201: Extract the buoy power consumption segment sequence based on the preset window length, calculate the total vibration acceleration at each sampling point, and denot the total vibration acceleration at the i-th sampling point as a. i a i =(a x 2 +a y 2 +a z 2 ) 1 / 2 Based on the total vibration acceleration analysis at each sampling point, the vibration energy index (EV) of the buoy body is analyzed. ; Where I represents the total number of sampling points in the preset window length; i∈[1,I], i is a positive integer; Analysis of buoy power consumption anomalies (AP) based on historical mean and standard deviation of buoy power consumption.t AP t =|P(t)-µ p | / σ p ; where µ p σ represents the historical average power consumption of the buoy. p The standard deviation of buoy power consumption; S202: Analyze the radial approach speed rs' = vs × cos(αs − θs) based on the ship's speed, course, and polar coordinates of its position relative to the buoy; Analyze the collision risk index PC based on the radial approach speed, PC = 1 / (1 + exp(-K1 × RC - K2 × ds / D)); where K1 and K2 represent sensitivity coefficients, which are system preset constants; exp() represents an exponential function with a natural number base; RC represents the ship approach rate, which is equal to the radial approach speed divided by the distance between the ship and the buoy. The collision risk index threshold is denoted as PC. th PC th =PC0×exp(-K3×Hs;where PC0 represents the basic threshold value, K3 represents the threshold coefficient, and the basic threshold value and the threshold coefficient are preset constants.

[0010] This application effectively distinguishes between normal wave impact and abnormal vibration modes such as structural resonance through triaxial acceleration synthesis calculation; adaptively identifies battery aging, sensor short circuits, and power system faults; radial approach velocity calculation avoids the limitations of the traditional fixed safety distance method and adapts to ship speed change and steering scenarios; nonlinear risks are converted into normalized probability values, and the threshold is dynamically adjusted with wave height to avoid false triggering caused by wind and wave interference. According to the above scheme, S3 includes the following: S301: Construct the environmental impact factor EB = β1 × |Tw - Tref| + β2 × S / S0; where β1 and β2 represent the preset factor coefficients; Tref represents the buoy's rated operating temperature; and S0 represents the reference salinity. Extract the vibration energy index, power consumption anomaly degree, and health degradation index at the last operating moment of the faulty buoy. The health degradation index is set by the administrator, generating a data point set {(EV1, AP1, HD1), (EV2, AP2, HD2), ..., (EV... j AP j HD j ), ..., (EV J AP J HD J )}, where EV j AP represents the vibration energy index of the j-th data point. j HD represents the power consumption anomaly degree of the j-th data point. jLet represent the health degradation index of the j-th data point, j∈[1,J], where j is a positive integer and J represents the total number of data points; after performing nonlinear fitting on the data points, a health degradation model is established: y=ω1×x1+ω2×x2+ω3, where x1 represents the independent variable corresponding to the vibration energy index divided by the maximum value of the vibration energy index, x2 represents the independent variable corresponding to the power consumption anomaly divided by the maximum value of the power consumption anomaly, y represents the dependent variable of the health degradation index, and ω1, ω2 and ω3 are all fitting coefficients; A risk value model is constructed based on the health deterioration index, collision risk index, and environmental impact factors: R = γ1 × HD + γ2 × PC + γ3 × EB; where γ1, γ2, and γ3 represent risk coefficients, which are preset constants; S302: If the risk value is greater than or equal to the first risk threshold or the health deterioration index is greater than the health deterioration index threshold and the environmental impact factor is greater than the environmental impact factor threshold, the risk level is Level 1; otherwise, if the risk value is greater than or equal to the second risk threshold, the risk level is Level 2; otherwise, if the risk value is greater than or equal to the third risk threshold, the risk level is Level 3; otherwise, it is normal.

[0011] This application uses the effect of salinity (S / S0) to reflect the accelerated effect of seawater corrosion on equipment life; temperature deviation (|Tw-Tref|) is related to the structural stress caused by the thermal expansion of the sealed chamber; the nonlinear relationship between vibration, power consumption and equipment health is fitted by historical fault data, breaking through the mechanical judgment of traditional threshold alarms; the first-level risk adopts "OR" logic to ensure priority handling in extreme situations (such as equipment failure and harsh environment); and the graded threshold avoids the system overreaction caused by sudden changes in a single indicator.

[0012] According to the above scheme, S4 includes the following: If the risk level is Level 1, send a device hibernation command and activate the backup power supply; if the risk level is Level 2, generate a ship collision avoidance path Path={Δα=arcsin(d safe / rs),Δv=vs−v safe}; where Δα represents the ship's collision avoidance deflection angle, d safe This represents the preset safety distance, Δv represents the ship's collision avoidance adjustment speed, and v safe Indicates the preset safe speed for the vessel; if the risk level is level three, activate the satellite emergency communication link and broadcast a distress message. When the false alarm rate exceeds the preset threshold, the data point set is expanded and the fitting coefficients are recalibrated.

[0013] This application uses false alarm rate monitoring to trigger model parameter recalibration and continuously optimizes the evaluation accuracy through incremental learning.

[0014] Another aspect of this application provides a safety monitoring system for large marine data buoys based on multi-source fusion. The system is applied to the above-mentioned safety monitoring method for large marine data buoys based on multi-source fusion. The system includes a buoy data acquisition module, an anomaly risk analysis module, a safety risk assessment module, and an early warning and response module. The buoy data acquisition module is used to synchronously collect buoy equipment operating parameters, marine environmental data, and surrounding vessel dynamic information through the buoy body sensor group, environmental monitoring unit, and ship identification module. The anomaly risk analysis module is used to extract time-frequency domain features and detect anomalies in the collected data, and to construct a dynamic assessment model for collision risk by combining it with ship trajectory prediction. The safety risk assessment module is used to establish an integrated assessment system for equipment health, environmental interference and ship threats, and to classify risk levels by combining real-time monitoring data and historical databases. The early warning and response module is used to generate early warning information based on risk level and automatically trigger equipment maintenance instructions and ship collision avoidance alarms.

[0015] According to the above scheme, the buoy data acquisition module includes a basic data acquisition unit and a data fusion unit; The basic data acquisition unit is used to acquire buoy body monitoring data, environmental monitoring data, and ship data; The data fusion unit is used to synchronize and align heterogeneous data in time; establish a spatial grid mapping model to convert the ship's position into polar coordinates relative to the buoy; and generate a fused dataset with timestamps.

[0016] According to the above scheme, the abnormal risk analysis module includes a power consumption abnormality analysis unit and a collision risk analysis unit; The power consumption anomaly analysis unit is used to extract the buoy power consumption segment sequence based on a preset window length, calculate the total vibration acceleration at each sampling point, analyze the vibration energy index of the buoy body based on the total vibration acceleration at each sampling point, and analyze the power consumption anomaly degree of the buoy based on the historical mean and standard deviation of the buoy power consumption. The collision risk analysis unit is used to analyze the radial approach speed based on the ship's speed, course, and polar coordinates of its position relative to the buoy; and to analyze collision risk indicators based on the radial approach speed.

[0017] According to the above scheme, the security risk assessment module includes a risk value model construction unit and a risk level classification unit; The risk value model construction unit is used to construct environmental impact factors, extract vibration energy indicators, power consumption anomalies, and health degradation indices at the last operating moment of the faulty buoy, and generate a set of data points; after performing nonlinear fitting on the data points, a health degradation model is established; and a risk value model is constructed based on the health degradation index, collision risk indicators, and environmental impact factors. If the risk value is greater than or equal to the first risk threshold, or the health deterioration index is greater than the health deterioration index threshold and the environmental impact factor is greater than the environmental impact factor threshold, the risk level is Level 1; otherwise, if the risk value is greater than or equal to the second risk threshold, the risk level is Level 2; otherwise, if the risk value is greater than or equal to the third risk threshold, the risk level is Level 3; otherwise, it is normal.

[0018] According to the above scheme, the early warning and handling module is used to send a device hibernation command and activate the backup power supply if the risk level is level one; generate a ship collision avoidance path if the risk level is level two; and activate the satellite emergency communication link and broadcast a distress message if the risk level is level three.

[0019] Compared with existing technologies, the beneficial effects of this invention are as follows: This application collects vibration, power consumption, and position drift parameters for combined monitoring, accurately capturing early signs of equipment mechanical failures, and synchronously resolving the phase difference problem between waves, ship motion, and equipment status; polar coordinate mapping converts the ship's position into a buoy relative motion model, simplifying the collision prediction calculation dimension; the timestamped fusion dataset provides standardized input for subsequent time-frequency analysis, avoiding conflicts in multi-source data formats; this application effectively distinguishes between normal wave impacts and abnormal vibration modes such as structural resonance through triaxial acceleration synthesis calculation; and it adaptively identifies battery aging, sensor short-circuit power system failures. The application addresses several key challenges: First, it employs radial approach speed calculation to overcome the limitations of traditional fixed safety distance methods, adapting to ship speed changes and steering scenarios. Second, it converts nonlinear risks into normalized probability values, dynamically adjusting thresholds with wave height to prevent false triggering caused by wind and waves. Third, it uses salinity to reflect the accelerated effect of seawater corrosion on equipment lifespan. Fourth, it correlates temperature deviation with structural stress caused by thermal expansion of the sealed chamber. Fifth, it uses historical fault data to fit the nonlinear relationship between vibration, power consumption, and equipment health, overcoming the mechanical judgment of traditional threshold alarms. Sixth, it uses "OR" logic for first-level risks to ensure priority handling of extreme situations. Seventh, it uses tiered thresholds to avoid overreactions caused by sudden changes in a single indicator. Finally, it monitors and recalibrates the trigger model parameters based on the false alarm rate, continuously optimizing evaluation accuracy through incremental learning. Attached Figure Description

[0020] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1This is a flowchart illustrating the safety monitoring method for large marine data buoys based on multi-source fusion according to the present invention. Figure 2 This is a schematic diagram of the structure of the large-scale marine data buoy safety monitoring system based on multi-source fusion according to the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Please see Figure 1 This invention provides a technical solution: a method for safety monitoring of large marine data buoys based on multi-source fusion, the method comprising the following steps: S1: Simultaneously collect buoy equipment operating parameters, marine environmental data, and dynamic information of surrounding vessels through the buoy body sensor group, environmental monitoring unit, and ship identification module; S1 includes the following: S101: Acquire buoy body monitoring data, including triaxial vibration acceleration a collected by a vibration sensor. x a y a z The system acquires buoy power consumption sequence P(t) via a current sensor, where t represents time; records buoy position drift ΔL via a GPS module; acquires environmental monitoring data, including wave height Hs and wave period TP obtained via wave radar; collects water temperature Tw, salinity S, and water depth D via a temperature, salinity, and depth meter (TDM) sensor; acquires ship data, including ship position, ship speed vs, and ship heading αs obtained by parsing AIS signals; and extracts ship draft ds through visual recognition. S102: Perform time synchronization and alignment on heterogeneous data; establish a spatial grid mapping model to convert ship positions into polar coordinates (rs, θs) relative to buoys; generate a timestamped fused dataset D={D b D e D v}, where D b For buoy body data, D e For environmental data, D v For ship data.

[0023] S2: Extract time-frequency domain features and detect anomalies from the collected data, and construct a dynamic assessment model for collision risk by combining ship trajectory prediction; S2 contains the following: S201: Extract the buoy power consumption segment sequence based on the preset window length, calculate the total vibration acceleration at each sampling point, and denot the total vibration acceleration at the i-th sampling point as a. i a i =(a x 2 +a y 2 +a z 2 ) 1 / 2 Based on the total vibration acceleration analysis at each sampling point, the vibration energy index (EV) of the buoy body is analyzed. ; Where I represents the total number of sampling points in the preset window length; i∈[1,I], i is a positive integer; Analysis of buoy power consumption anomalies (AP) based on historical mean and standard deviation of buoy power consumption. t AP t =|P(t)-µ p | / σ p ; where µ p σ represents the historical average power consumption of the buoy. p The standard deviation of buoy power consumption; S202: Analyze the radial approach speed rs' = vs × cos(αs − θs) based on the ship's speed, course, and polar coordinates relative to the buoy; Analyze the collision risk index PC based on the radial approach speed, PC = 1 / (1 + exp(-K1 × RC - K2 × ds / D)); where K1 and K2 represent sensitivity coefficients, which are preset constants of the system; RC represents the ship approach rate, which is equal to the radial approach speed divided by the distance between the ship and the buoy. The collision risk index threshold is denoted as PC. th PC th =PC0×exp(-K3×Hs;where PC0 represents the basic threshold value, K3 represents the threshold coefficient, and the basic threshold value and threshold coefficient are preset constants.

[0024] S3: Establish a fusion assessment system for equipment health, environmental interference, and ship threats, and classify risk levels by combining real-time monitoring data and historical databases; S3 includes the following: S301: Construct the environmental impact factor EB = β1 × |Tw - Tref| + β2 × S / S0; where β1 and β2 represent the preset factor coefficients; Tref represents the buoy's rated operating temperature; and S0 represents the reference salinity. Extract the vibration energy index, power consumption anomaly degree, and health degradation index from the last operating moment of the faulty buoy. The health degradation index is set by the administrator, generating a data point set {(EV1, AP1, HD1), (EV2, AP2, HD2), ..., (EV... j AP j HD j ), ..., (EV J AP J HD J )}, where EV j AP represents the vibration energy index of the j-th data point. j HD represents the power consumption anomaly degree of the j-th data point. j Let represent the health degradation index of the j-th data point, j∈[1,J], where j is a positive integer and J represents the total number of data points; after performing nonlinear fitting on the data points, a health degradation model is established: y=ω1×x1+ω2×x2+ω3, where x1 represents the independent variable corresponding to the vibration energy index divided by the maximum value of the vibration energy index, x2 represents the independent variable corresponding to the power consumption anomaly divided by the maximum value of the power consumption anomaly, y represents the dependent variable of the health degradation index, and ω1, ω2 and ω3 are all fitting coefficients; A risk value model is constructed based on the health deterioration index, collision risk indicators, and environmental impact factors: R = γ1 × HD + γ2 × PC + γ3 × EB; where γ1, γ2, and γ3 represent risk coefficients, which are preset constants; S302: If the risk value is greater than or equal to the first risk threshold or the health deterioration index is greater than the health deterioration index threshold and the environmental impact factor is greater than the environmental impact factor threshold, the risk level is Level 1; otherwise, if the risk value is greater than or equal to the second risk threshold, the risk level is Level 2; otherwise, if the risk value is greater than or equal to the third risk threshold, the risk level is Level 3; otherwise, it is normal.

[0025] Example 1: In this example, HD=0.82, PC=0.3, EB=0.4; γ1=0.6, γ2=0.2, γ3=0.2; Therefore, R = 0.6 × 0.8² + 0.2 × 0.3 + 0.2 × 0.4 = 0.6²; In this embodiment, the first threshold for risk value is 0.8, and the threshold for health degradation index is 0.75; although R=0.62<0.8, Hd>0.75, triggering a Level I warning (equipment failure takes priority).

[0026] S4: Generates early warning information based on risk level, automatically triggering equipment maintenance instructions and ship collision avoidance alarms.

[0027] S4 includes the following: If the risk level is Level 1, send a device hibernation command and activate the backup power supply; if the risk level is Level 2, generate a ship collision avoidance path Path={Δα=arcsin(d safe / rs),Δv=vs−v safe}; where Δα represents the ship's collision avoidance deflection angle, d safe This represents the preset safety distance, Δv represents the ship's collision avoidance adjustment speed, and v safe Indicates the preset safe speed for the vessel; if the risk level is level three, activate the satellite emergency communication link and broadcast a distress message. When the false alarm rate exceeds the preset threshold, the data point set is expanded and the fitting coefficients are recalibrated.

[0028] Please see Figure 2 The present invention provides a technical solution: a large-scale marine data buoy safety monitoring system based on multi-source fusion, which includes a buoy data acquisition module, an anomaly risk analysis module, a safety risk assessment module, and an early warning and response module; The buoy data acquisition module is used to synchronously collect buoy equipment operating parameters, marine environmental data, and dynamic information of surrounding vessels through the buoy body sensor group, environmental monitoring unit, and ship identification module. The anomaly risk analysis module is used to extract time-frequency domain features and detect anomalies in the collected data, and to build a dynamic collision risk assessment model in combination with ship trajectory prediction. The safety risk assessment module is used to establish an integrated assessment system for equipment health, environmental interference, and ship threats, and to classify risk levels by combining real-time monitoring data and historical databases; The early warning and response module is used to generate early warning information based on risk level and automatically trigger equipment maintenance instructions and ship collision avoidance alarms.

[0029] The buoy data acquisition module includes a basic data acquisition unit and a data fusion unit; The basic data acquisition unit is used to acquire buoy body monitoring data, environmental monitoring data, and ship data; The data fusion unit is used to synchronize and align heterogeneous data in time; establish a spatial grid mapping model to convert the ship's position into polar coordinates relative to the buoy; and generate a fused dataset with timestamps.

[0030] The anomaly risk analysis module includes a power consumption anomaly analysis unit and a collision risk analysis unit; The power consumption anomaly analysis unit is used to extract the buoy power consumption segment sequence based on a preset window length, calculate the total vibration acceleration at each sampling point, analyze the vibration energy index of the buoy body based on the total vibration acceleration at each sampling point, and analyze the power consumption anomaly of the buoy based on the historical mean and standard deviation of the buoy power consumption. The collision risk analysis unit is used to analyze radial approach speed based on the ship's speed, course, and polar coordinates of its position relative to the buoy; and to analyze collision risk indicators based on radial approach speed.

[0031] The safety risk assessment module includes a risk value model construction unit and a risk level classification unit; The risk value model building unit is used to construct environmental impact factors, extract vibration energy indicators, power consumption anomaly degree, and health degradation index at the last operating moment of the faulty buoy, and generate a set of data points; after nonlinear fitting of the data points, a health degradation model is established; and a risk value model is constructed based on the health degradation index, collision risk indicators, and environmental impact factors. If the risk value is greater than or equal to the first risk threshold, or the health deterioration index is greater than the health deterioration index threshold and the environmental impact factor is greater than the environmental impact factor threshold, the risk level is Level 1; otherwise, if the risk value is greater than or equal to the second risk threshold, the risk level is Level 2; otherwise, if the risk value is greater than or equal to the third risk threshold, the risk level is Level 3; otherwise, it is normal.

[0032] The early warning and response module is used to send a device hibernation command and activate the backup power supply if the risk level is Level 1; generate a collision avoidance path for the ship if the risk level is Level 2; and activate the satellite emergency communication link and broadcast a distress message if the risk level is Level 3.

[0033] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0034] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for safety monitoring of large marine data buoys based on multi-source fusion, characterized in that, The method includes the following steps: S1: Simultaneously collect buoy equipment operating parameters, marine environmental data, and dynamic information of surrounding vessels through the buoy body sensor group, environmental monitoring unit, and ship identification module; S1 includes the following: S101: Acquire buoy body monitoring data, the buoy body monitoring data including triaxial vibration acceleration a collected by a vibration sensor. x a y a z The system acquires the buoy power consumption sequence P(t) via a current sensor, where t represents time; records the buoy position drift ΔL via a GPS module; acquires environmental monitoring data, including wave height Hs and wave period TP obtained via wave radar; collects water temperature Tw, salinity S, and water depth D via a temperature, salinity, and depth meter sensor; and acquires ship data, including ship position, ship speed vs, and ship heading αs obtained by parsing AIS signals. The ship's draft (ds) is extracted using visual recognition. S102: Perform time synchronization and alignment on heterogeneous data; establish a spatial grid mapping model to convert ship positions into polar coordinates (rs, θs) relative to buoys; generate a timestamped fused dataset D={D b D e D v }, where D b For buoy body data, D e For environmental data, D v For ship data; S2: Extract time-frequency domain features and detect anomalies from the collected data, and construct a dynamic assessment model for collision risk by combining ship trajectory prediction; S2 contains the following: S201: Extract the buoy power consumption segment sequence based on the preset window length, calculate the total vibration acceleration at each sampling point, and denot the total vibration acceleration at the i-th sampling point as a. i a i =(a x 2 +a y 2 +a z 2 ) 1 / 2 The vibration energy index (EV) of the buoy body is analyzed based on the total vibration acceleration at each sampling point. Analysis of buoy power consumption anomalies (AP) based on historical mean and standard deviation of buoy power consumption. t AP t =|P(t)-µ p | / σ p ; where µ p σ represents the historical average power consumption of the buoy. p The standard deviation of buoy power consumption; S202: Analyze the radial approach speed rs' = vs × cos(αs − θs) based on the ship's speed, course, and polar coordinates of its position relative to the buoy; Analyze the collision risk index PC based on the radial approach speed, PC = 1 / (1 + exp(-K1 × RC - K2 × ds / D)); where K1 and K2 represent sensitivity coefficients, which are preset constants of the system, exp() represents an exponential function with natural numbers as the base; RC represents the ship approach rate, which is equal to the radial approach speed divided by the distance between the ship and the buoy. The collision risk index threshold is denoted as PC. th PC th =PC0×exp(-K3×Hs;where PC0 represents the basic threshold value, K3 represents the threshold coefficient, and the basic threshold value and the threshold coefficient are preset constants; S3: Establish a fusion assessment system for equipment health, environmental interference, and ship threats, and classify risk levels by combining real-time monitoring data and historical databases; S4: Generates early warning information based on risk level, automatically triggering equipment maintenance instructions and ship collision avoidance alarms.

2. The method for safety monitoring of large marine data buoys based on multi-source fusion as described in claim 1, characterized in that: S3 includes the following: S301: Construct the environmental impact factor EB = β1 × |Tw - Tref| + β2 × S / S0; where β1 and β2 represent the preset factor coefficients; Tref represents the buoy's rated operating temperature; and S0 represents the reference salinity. Extract the vibration energy index, power consumption anomaly degree, and health degradation index at the last operating moment of the faulty buoy. The health degradation index is set by the administrator, generating a data point set {(EV1, AP1, HD1), (EV2, AP2, HD2), ..., (EV... j AP j HD j ), ..., (EV J AP J HD J )}, where EV j AP represents the vibration energy index of the j-th data point. j HD represents the power consumption anomaly degree of the j-th data point. j Let represent the health degradation index of the j-th data point, j∈[1,J], where j is a positive integer and J represents the total number of data points; after performing nonlinear fitting on the data points, a health degradation model is established: y=ω1×x1+ω2×x2+ω3, where x1 represents the independent variable corresponding to the vibration energy index divided by the maximum value of the vibration energy index, x2 represents the independent variable corresponding to the power consumption anomaly divided by the maximum value of the power consumption anomaly, y represents the dependent variable of the health degradation index, and ω1, ω2 and ω3 are all fitting coefficients; A risk value model is constructed based on the health deterioration index, collision risk index, and environmental impact factors: R = γ1 × HD + γ2 × PC + γ3 × EB; where γ1, γ2, and γ3 represent risk coefficients, which are preset constants; S302: If the risk value is greater than or equal to the first risk threshold or the health deterioration index is greater than the health deterioration index threshold and the environmental impact factor is greater than the environmental impact factor threshold, the risk level is Level 1; otherwise, if the risk value is greater than or equal to the second risk threshold, the risk level is Level 2; otherwise, if the risk value is greater than or equal to the third risk threshold, the risk level is Level 3; otherwise, it is normal.

3. The method for safety monitoring of large marine data buoys based on multi-source fusion according to claim 2, characterized in that: S4 includes the following: If the risk level is Level 1, send a device hibernation command and activate the backup power supply; if the risk level is Level 2, generate a ship collision avoidance path Path={Δα=arcsin(d safe / rs),Δv=vs−v safe }; where Δα represents the ship's collision avoidance deflection angle, d safe This represents the preset safety distance, Δv represents the ship's collision avoidance adjustment speed, and v safe Indicates the preset safe speed for the vessel; if the risk level is level three, activate the satellite emergency communication link and broadcast a distress message. When the false alarm rate exceeds the preset threshold, the data point set is expanded and the fitting coefficients are recalibrated.

4. A large-scale marine data buoy safety monitoring system based on multi-source fusion, wherein the system is applied to the large-scale marine data buoy safety monitoring method based on multi-source fusion as described in any one of claims 1-3, characterized in that, The system includes a buoy data acquisition module, an anomaly risk analysis module, a safety risk assessment module, and an early warning and response module; The buoy data acquisition module is used to synchronously collect buoy equipment operating parameters, marine environmental data, and surrounding vessel dynamic information through the buoy body sensor group, environmental monitoring unit, and ship identification module. The anomaly risk analysis module is used to extract time-frequency domain features and detect anomalies in the collected data, and to construct a dynamic assessment model for collision risk by combining it with ship trajectory prediction. The safety risk assessment module is used to establish an integrated assessment system for equipment health, environmental interference and ship threats, and to classify risk levels by combining real-time monitoring data and historical databases. The early warning and response module is used to generate early warning information based on risk level and automatically trigger equipment maintenance instructions and ship collision avoidance alarms.

5. The large-scale marine data buoy safety monitoring system based on multi-source fusion as described in claim 4, characterized in that: The buoy data acquisition module includes a basic data acquisition unit and a data fusion unit; The basic data acquisition unit is used to acquire buoy body monitoring data, environmental monitoring data, and ship data; The data fusion unit is used to perform time synchronization and alignment of heterogeneous data. A spatial grid mapping model is established to convert the ship's position into polar coordinates relative to the buoy; a fused dataset with timestamps is generated.

6. The large-scale marine data buoy safety monitoring system based on multi-source fusion according to claim 4, characterized in that: The anomaly risk analysis module includes a power consumption anomaly analysis unit and a collision risk analysis unit; The power consumption anomaly analysis unit is used to extract the buoy power consumption segment sequence based on a preset window length, calculate the total vibration acceleration at each sampling point, analyze the vibration energy index of the buoy body based on the total vibration acceleration at each sampling point, and analyze the power consumption anomaly degree of the buoy based on the historical mean and standard deviation of the buoy power consumption. The collision risk analysis unit is used to analyze the radial approach speed based on the ship's speed, course, and polar coordinates of its position relative to the buoy. Collision risk indicators are analyzed based on radial approach velocity.

7. The large-scale marine data buoy safety monitoring system based on multi-source fusion as described in claim 4, characterized in that: The security risk assessment module includes a risk value model construction unit and a risk level classification unit; The risk value model construction unit is used to construct environmental impact factors, extract vibration energy indicators, power consumption anomaly degree, and health degradation degree index at the last operating moment of the faulty buoy, and generate a set of data points; after performing nonlinear fitting on the data points, a health degradation degree model is established. A risk value model was constructed based on the health degradation index, collision risk indicators, and environmental impact factors. If the risk value of the risk level classification unit is greater than or equal to the first threshold of the risk value, or the health deterioration index is greater than the health deterioration index threshold and the environmental impact factor is greater than the environmental impact factor threshold, the risk level is level one. Otherwise, if the risk value is greater than or equal to the second risk threshold, the risk level is level two; otherwise, if the risk value is greater than or equal to the third risk threshold, the risk level is level three; otherwise, it is normal.

8. The large-scale marine data buoy safety monitoring system based on multi-source fusion according to claim 4, characterized in that: The early warning and response module is used to send a device hibernation command and activate the backup power supply if the risk level is Level 1; generate a collision avoidance path for the ship if the risk level is Level 2; and activate the satellite emergency communication link and broadcast a distress message if the risk level is Level 3.

Images