An unmanned ship oil spill accident intelligent sensing system

Abstract

This invention relates to the field of maritime traffic control technology and discloses an intelligent perception system for unmanned vessel oil spill accidents, comprising: a tidal current analysis unit, an oil film analysis unit, a traffic analysis unit, and a traffic control unit. The tidal current analysis unit generates a tidal dynamic field containing navigability, tidal hazard intensity, etc., thereby reflecting the risk of the flow field. The oil film analysis unit collaboratively analyzes the oil spill data with the tidal dynamic field to generate an oil film disturbance motion vector, reflecting channel encroachment and navigation obstruction. The traffic analysis unit generates a predicted hazardous spatiotemporal set containing hazardous time periods, areas, and levels to cover the potential drift range of the oil film. The traffic control unit plots the dynamically predicted hazardous area and issues traffic control commands, thereby achieving accurate prediction and dynamic control of oil film drift, avoiding traffic chaos in the main channel, improving the accuracy of traffic control in strong tidal port channels, and ensuring navigation safety.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of maritime traffic control, in particular to an intelligent perception system for oil spill accidents of unmanned ships. BACKGROUND

[0002] Currently, when an oil spill accident occurs, sensors carried by unmanned ships are usually used to perceive the oil film in real time, and based on the current position and range thereof, the oil film is marked as a dangerous area on a maritime electronic chart, and traffic in the dangerous area is regulated or surrounding ships are guided to achieve the effect of maritime traffic control.

[0003] However, the above control method still has the following defects when applied in narrow channels of ports affected by strong tides, specifically: although the existing unmanned ships can divide a dangerous area after perceiving an oil spill accident, due to the change in the flow direction of the tide, the oil film may drift downstream to the main channel as a whole at the next tide. This drift is random, and although the unmanned ship can perceive the oil spill situation in real time and report the instantaneous position of the oil film, the dangerous area divided cannot predict the change of the oil film in the tide, resulting in that many ships still adjust the channel according to the original dangerous area, which is easy to cause chaos in maritime traffic in the main channel, and the control effect of maritime traffic is poor. SUMMARY

[0004] In view of the deficiencies of the prior art, the present application provides an intelligent perception system for oil spill accidents of unmanned ships, which solves the above problems.

[0005] The above technical purpose of the present application is achieved by the following technical scheme:

[0006] An intelligent perception system for oil spill accidents of unmanned ships, comprising:

[0007] A tidal current analysis unit configured to acquire grid-based tidal current forecast data of a target channel for at least two future tidal periods from a maritime data center, analyze the tidal current forecast data, and obtain a tidal dynamic field, the target channel being a port channel;

[0008] An oil film analysis unit configured to acquire oil spill data perceived by the unmanned ship in real time, and cooperatively analyze an oil film center point in the oil spill data with the tidal dynamic field to generate an oil film disturbed motion vector;

[0009] A traffic analysis unit configured to take the oil film center point as a tracking starting point, analyze a drift path of the oil film based on the oil film disturbed motion vector and the tidal dynamic field, and generate a predicted dangerous spatio-temporal set;

[0010] A traffic control unit configured to draw a dynamic predicted dangerous area according to the predicted dangerous spatio-temporal set, and issue a traffic control instruction.

[0011] Furthermore, the tidal current forecast data is analyzed to obtain the tidal dynamic field, including:

[0012] Analyze the tidal current forecast data and calculate the accessibility of each grid node at different time points;

[0013] The intensity of tidal current hazard is obtained by analyzing the abrupt changes in flow direction angle and velocity difference between adjacent grid nodes;

[0014] By analyzing the vortex intensity of each grid node and combining the vortex intensity with the flow velocity from the tidal current forecast data, vortex-induced hazard energy that poses a potential threat to ship stability is generated.

[0015] Furthermore, analysis of tidal current forecast data yields the tidal dynamic field, which also includes:

[0016] By integrating accessibility, tidal current hazard intensity, and vortex hazard energy, a comprehensive traffic impact value is generated for each grid node.

[0017] Based on the comprehensive traffic impact value, the degree of traffic impact is assessed, and a tidal dynamic field is generated.

[0018] Furthermore, the oil film center point in the oil spill data is analyzed in conjunction with the tidal dynamic field to generate the disturbed motion vector of the oil film, including:

[0019] Obtain the navigation area of ​​the target channel;

[0020] Based on the center point of the oil film, the overlapping area between the current contour boundary of the oil film and the navigation area is analyzed to generate the channel occupancy rate;

[0021] The navigation resistance potential energy value is obtained by calculating the channel occupancy rate and oil film thickness.

[0022] Furthermore, by jointly analyzing the oil film center point in the oil spill data with the tidal dynamic field, a disturbed motion vector of the oil film is generated, which also includes:

[0023] Analyze the navigation resistance potential energy and tidal dynamic field, and calculate the dynamic hazard coupling factor of each grid node;

[0024] The coupling factor between the navigation resistance potential energy value and the dynamic hazard is calculated, and the transmission direction and intensity of the oil film along the target channel under the action of tidal current are analyzed to generate the oil film threat transmission vector.

[0025] Furthermore, by jointly analyzing the oil film center point in the oil spill data with the tidal dynamic field, a disturbed motion vector of the oil film is generated, which also includes:

[0026] Identify key route nodes in the target channel, analyze the impact of oil slick threat transmission vectors on key route nodes, and generate risk focus of key route nodes;

[0027] The oil film threat transmission vector and risk focus are fused to generate the oil film disturbance motion vector.

[0028] Furthermore, taking the center point of the oil film as the starting point for tracking, and based on the disturbed motion vector of the oil film and the tidal dynamic field, the drift path of the oil film is analyzed to generate a predicted hazardous spatiotemporal set, including:

[0029] Based on the disturbed motion vector of the oil film, a multi-directional dynamic analysis is performed starting from the center point of the oil film in the tidal dynamic field to generate a path cell sequence.

[0030] For each path cell in the path cell sequence, the tidal dynamic field and the navigation resistance potential energy value are fused to generate the drift stress intensity representing oil film deformation and migration.

[0031] Based on the path cell sequence and drift stress intensity, the state of the oil film during the motion process is analyzed, and the spatiotemporal diffusion entropy of the oil film is generated.

[0032] Furthermore, taking the center point of the oil film as the starting point for tracking, and based on the disturbed motion vector of the oil film and the tidal dynamic field, the drift path of the oil film is analyzed to generate a predicted hazardous spatiotemporal set, which also includes:

[0033] Based on the path cell sequence, the drift stress intensity and oil film spatiotemporal diffusion entropy are analyzed to generate dynamic risk loads;

[0034] The dynamic risk load is analyzed to extract areas where the risk value continuously exceeds the preset critical value and has spatiotemporal continuity, generating a predicted spatiotemporal set of hazards that includes dangerous periods, dangerous areas and dangerous levels.

[0035] Furthermore, based on the predicted hazard spatiotemporal set, a dynamic predicted hazard zone is drawn, and traffic control instructions are issued, including:

[0036] By analyzing the predicted hazard spatiotemporal set, the hazard zone weight coefficients are obtained;

[0037] By matching the weight coefficient of the danger zone with the tidal dynamic field, a dynamic boundary correction factor is generated.

[0038] Furthermore, based on the predicted hazard spatiotemporal set, a dynamic predicted hazard zone is drawn, and traffic control instructions are issued, which also includes:

[0039] Based on the dynamic boundary correction factor and the hazard zone weight coefficient, the dynamic predicted hazard zone is plotted to evolve over time.

[0040] Traffic control instructions are issued based on dynamic predictions of danger zones.

[0041] In summary, the present invention has the following main beneficial effects:

[0042] The tidal current analysis unit acquires gridded tidal current forecast data for at least two future tidal cycles, generating a tidal dynamic field that includes navigability, tidal current hazard intensity, vortex hazard energy, and comprehensive traffic impact value. This determines the risk characteristics of the port channel flow field. The oil film analysis unit performs collaborative analysis of oil spill data sensed by unmanned vessels with the tidal dynamic field. Through layer-by-layer calculations of channel occupancy rate, navigation obstruction potential energy value, dynamic hazard coupling factor, oil film threat transmission vector, and risk focus, it generates an oil film disturbance motion vector, thereby reflecting the degree of oil film encroachment on the channel and navigation obstruction effect.

[0043] Starting from the center point of the oil slick, the traffic analysis unit generates a predicted hazardous spatiotemporal set containing hazardous periods, hazardous areas, and hazardous levels based on the analysis of path cell sequence, drift stress intensity, spatiotemporal diffusion entropy of the oil slick, and dynamic risk load. This comprehensively covers the potential drift range of the oil slick. The traffic control unit, based on the hazardous area weight coefficient and dynamic boundary correction factor, plots the dynamically predicted hazardous area that evolves over time. The drift trajectory of the oil slick is visually presented in the next 24 hours through graded warning color bands, and precise traffic control commands are issued. This achieves accurate prediction and dynamic control of oil slick drift, avoids traffic chaos in the main channel caused by ships adjusting their channels based on static hazardous areas, improves the accuracy and effectiveness of traffic control in narrow channels of strong tidal ports, and ensures navigation order and maritime traffic safety. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of an intelligent sensing system for oil spill accidents from an unmanned vessel, according to the present invention. Detailed Implementation

[0045] 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.

[0046] refer to Figure 1 An intelligent sensing system for unmanned vessel oil spill accidents includes:

[0047] The tidal current analysis unit is used to acquire gridded tidal current forecast data of the target channel in real time from the maritime data center for at least two tidal cycles in the future, analyze the tidal current forecast data to obtain the tidal dynamic field, wherein the grid of the tidal current forecast data is 100 meters × 100 meters, and two tidal cycles are 24 hours, and the target channel is the narrow channel of the port.

[0048] The tidal current forecast data includes the tidal current (ocean current) velocity and direction angle at each grid node within the target channel at each time point;

[0049] The oil film analysis unit is used to acquire oil spill data sensed by the unmanned vessel in real time, and to perform collaborative analysis of the oil film center point in the oil spill data with the tidal dynamic field to generate the disturbed motion vector of the oil film.

[0050] The oil spill data includes: the center point of the oil film, the current outline boundary of the oil film, and the thickness of the oil film;

[0051] The traffic analysis unit is used to analyze the drift path of the oil film, taking the center point of the oil film as the starting point of tracking, and based on the disturbed motion vector of the oil film and the tidal dynamic field, to generate a predicted dangerous spatiotemporal set.

[0052] Traffic control units are used to map dynamic predicted hazard zones based on predicted spatiotemporal sets of hazards and issue traffic control commands.

[0053] In one embodiment, the tidal current forecast data is analyzed to obtain the tidal dynamic field, including:

[0054] The tidal current forecast data is analyzed to calculate the passability of each grid node at different time points. Specifically, for each grid node, the velocity sequence and flow direction angle sequence are calculated at 15-minute intervals for the next 24 hours. The ratio of the velocity standard deviation to the average velocity is calculated as the velocity variation coefficient, and the standard deviation of the flow direction angle sequence is calculated as the flow direction stability index.

[0055] The velocity variation coefficient is multiplied by the flow direction stability index to obtain the comprehensive index. The velocity threshold is set to 0.5 m / s. When the predicted velocity of a grid node at a certain time point is less than or equal to the velocity threshold, the navigability is 1. When the predicted velocity of a grid node at a certain time point is greater than the velocity threshold, 1 is subtracted (the predicted velocity minus the velocity threshold is divided by the difference between the maximum predicted velocity of all grid nodes and the velocity threshold), and the result is multiplied by the comprehensive index and normalized to the 0-1 range. This gives the navigability of each grid at different time points. Navigability is mainly used to reflect the navigation safety of grid nodes and indicates the navigation difficulty of ships.

[0056] The intensity of tidal current hazard is obtained by analyzing the abrupt changes in flow direction angles and velocity differences between adjacent grid nodes. Specifically, this includes: for each grid node at a given time point, obtaining the flow direction angles and velocity data of its four adjacent grid nodes (east, south, west, and north); calculating the absolute angle difference between the flow direction angle of the grid node and the flow direction angles of each adjacent grid node, and taking the maximum value as the abrupt change value of the flow direction; and simultaneously calculating the absolute difference between the flow velocity of the grid node and the flow velocity of each adjacent grid node, and taking the maximum value as the velocity difference.

[0057] The flow direction change value and the velocity difference value are normalized to the 0-1 range respectively. The weight of the flow direction change value is set to 0.7 and the weight of the velocity difference value is set to 0.3. The normalized flow direction change value and velocity difference value are multiplied by their corresponding weights and then added together to obtain the tidal current hazard intensity of the grid node at that time point. The tidal current hazard intensity is mainly used to identify the flow field shear zone and vortex-prone zone. These areas can cause ships to lose course control and are key areas to avoid in traffic control.

[0058] Among them, the flow direction change value mainly represents the drastic change in flow direction between adjacent grids. Because it will directly form local flow field shear and vortex structure, it can not only cause the ship's rudder effect to fail and the course to be out of control, but also cause the oil film to be torn and stretched rapidly, increasing the difficulty of diffusion prediction. Therefore, the weight is relatively high at 0.7. The velocity difference value represents the velocity difference between adjacent grids. Although it will also affect navigation energy consumption and oil film deformation, its harm is usually gradual and linear. Therefore, the weight is relatively low at 0.3.

[0059] The vortex intensity of each grid node is analyzed, and the vortex intensity is combined with the flow velocity of the tidal current forecast data to generate vortex hazard energy that poses a potential threat to ship stability. Specifically, based on a grid spacing of 100 meters, for the central grid node, the northward flow velocity components of its east and west adjacent grid nodes are obtained, and the difference between the two is calculated to obtain the northward flow velocity component difference value; at the same time, the eastward flow velocity components of its south and north adjacent grid nodes are obtained, and the difference between the two is calculated to obtain the eastward flow velocity component difference value.

[0060] Divide the difference in the northward velocity component by 200 meters, then divide the difference in the eastward velocity component by 200 meters, and then subtract the two results to obtain the vertical vorticity value of the grid; multiply the absolute value of the vertical vorticity value by 100 to obtain the vortex intensity.

[0061] The square of the eastward velocity component is added to the square of the northward velocity component, and the square root of the sum is taken to obtain the resultant velocity value of the grid node. The resultant velocity value represents the actual speed of the water flow at the grid node.

[0062] Multiplying the vortex intensity by the resultant velocity value yields the vortex hazard energy of the grid node at that time point. The vortex hazard energy represents the rotational intensity and flow kinetic energy of the vortex, reflecting its potential threat to ship stability.

[0063] In one embodiment, analyzing tidal current forecast data to obtain the tidal dynamic field further includes:

[0064] By integrating accessibility, tidal current hazard intensity, and vortex hazard energy, a comprehensive traffic impact value for each grid node is generated. Specifically, this involves: normalizing the tidal current hazard intensity and vortex hazard energy to the 0-1 interval; multiplying the normalized tidal current hazard intensity and vortex hazard energy by accessibility; adding the two multiplication results; performing a cube root operation on the sum; and normalizing the result to the 0-1 interval to obtain the comprehensive traffic impact value for that grid node. The comprehensive traffic impact value serves as a core indicator of the tidal dynamic field, enabling the consideration of oil film location and flow field risk during subsequent hazard zone delineation, thereby improving the accuracy of traffic control.

[0065] Based on the comprehensive traffic impact value, the degree of traffic impact is assessed, and a tidal dynamic field is generated. Specifically, for each grid node, the maximum value of the comprehensive traffic impact value for its next eight consecutive time points is extracted, where the next eight consecutive time points represent the next two hours.

[0066] Calculate the ratio of the comprehensive traffic impact value of the grid node at the current time point to the average comprehensive traffic impact value of all grid nodes in the entire target waterway to obtain its relative spatial risk intensity; then divide the number of time points in the past hour where the comprehensive traffic impact value of the grid node exceeds 0.5 by four to obtain the proportion of high-risk duration.

[0067] The traffic impact index of the grid node is obtained by multiplying the maximum comprehensive traffic impact value, the relative intensity of spatial risk, and the duration ratio, and then normalizing the result to the 0-1 range; the tidal flow forecast data, oil spill data, and traffic impact index are combined.

[0068] For each grid node at each time point, a unique data structure layer is generated. The first layer of the data structure layer is: tidal current forecast data; the second layer of the data structure layer is: oil spill data; and the third layer of the data structure layer is: traffic impact index. This yields a structured tidal dynamic field.

[0069] By deeply analyzing tidal current forecast data, a structured tidal dynamic field is generated, which includes navigability, tidal current hazard intensity, vortex hazard energy, and comprehensive traffic impact value. This accurately captures the changing patterns of tidal flow and the drift trend of oil films, enabling the prediction of the future location of oil films. It solves the problem that existing danger zone delineation cannot adapt to tidal influences, and can dynamically guide the updating of danger zones. This avoids traffic chaos in the main channel caused by ships adjusting the channel based on static danger zones, and improves the accuracy of maritime traffic control in narrow channels of ports under strong tidal influences.

[0070] In one embodiment, the oil film center point in the oil spill data is analyzed in conjunction with the tidal dynamic field to generate a disturbed motion vector of the oil film, including:

[0071] Obtain the navigation area of ​​the target channel;

[0072] Based on the center point of the oil film, the overlapping area between the current contour boundary of the oil film and the navigation area is analyzed to generate the channel occupancy rate. Specifically, this includes: obtaining the main channel axis of the navigation area; drawing a perpendicular line from the center point of the oil film to the main channel axis to obtain the foot point of the perpendicular; extending 100 meters upstream and downstream along the main channel axis with the foot point as the center, and taking the 200-meter axis segment as the analysis section; and taking the intersection area of ​​the oil film contour boundary and the analysis section as the oil film intrusion area.

[0073] Obtain the outline area of ​​the oil film; calculate the ratio of the oil film intrusion area to the outline area of ​​the oil film to obtain the area intrusion rate; simultaneously calculate the ratio of the projected length of the oil film intrusion area along the main channel axis to the total analysis length of 200 meters to obtain the length blocking rate; multiply the area intrusion rate and the length blocking rate to obtain the channel occupancy rate of the analysis section. The channel occupancy rate mainly reflects the proportion of the oil film encroaching on the main channel, thereby judging the degree of decline in channel capacity.

[0074] The navigation resistance potential energy value is obtained by calculating the channel occupancy rate and oil film thickness. Specifically, this includes obtaining the oil film thickness in all historical oil spill accident records of the target channel in the past year, calculating the average of all historical oil film thicknesses, and using it as a reference thickness threshold.

[0075] Calculate the ratio of the average oil film thickness in the current oil film invasion zone to the reference thickness threshold to obtain the thickness influence coefficient. If the thickness influence coefficient is greater than 1, the thickness influence coefficient is 1; otherwise, the original value is maintained.

[0076] Obtain the time from the occurrence of this oil spill to the present moment, in hours. Divide the time by 24 to obtain the time decay factor.

[0077] The target channel is re-divided into multiple uniform fixed sections at 200-meter intervals along the main channel axis. The channel occupancy rate, thickness influence coefficient, and time decay factor are then multiplied to obtain the navigation resistance potential energy value of the fixed section at the current moment. The navigation resistance potential energy value is used to reflect the degree of resistance of the oil film to navigation.

[0078] In one embodiment, the oil film center point in the oil spill data is analyzed in conjunction with the tidal dynamic field to generate a disturbed motion vector of the oil film, and the method further includes:

[0079] Analyze the navigation resistance potential energy value and tidal dynamic field, and calculate the dynamic hazard coupling factor of each grid node. Specifically, this includes: associating each grid node with its fixed section; calculating the 75th percentile of the navigation resistance potential energy value of all fixed sections in this oil spill accident, and using it as the blockage threshold.

[0080] If the navigation resistance potential energy value of the fixed section to which the grid node belongs is less than the blockage threshold, it indicates that its oil film blockage risk is relatively low, and its dynamic risk coupling factor is 0.

[0081] If the navigation obstruction potential energy value of the fixed section to which the grid node belongs is greater than or equal to the blockage threshold, it indicates that the risk of oil film blockage is relatively high. Then, calculate the ratio of the traffic impact index in the tidal dynamic field corresponding to the grid node to the average traffic impact index of all grid nodes in the entire target channel, and subtract 1 from the ratio to obtain the relative risk intensity.

[0082] Multiply the navigation obstruction potential energy value, relative risk intensity, and traffic impact index of the fixed section to which the grid node belongs, and normalize the product to the 0-1 interval to obtain the dynamic risk coupling factor of the grid node. The dynamic risk coupling factor is used to identify high-risk grids with severe oil film blockage and high flow field risk.

[0083] The coupling factor between the navigation resistance potential energy value and the dynamic hazard is calculated, and the transmission direction and intensity of the oil film along the target channel under the action of tidal current are analyzed to generate the oil film threat transmission vector. Specifically, this includes: obtaining the fixed section where the center point of the oil film is located and taking this fixed section as the center section.

[0084] Find two adjacent fixed sections upstream and downstream of the central section along the main channel axis; for these five sections, multiply the navigation obstruction potential energy value of each section by the mean of the dynamic hazard coupling factor of all grid nodes in that section to obtain the integrated threat value of each section.

[0085] The current tidal flow angle of the central grid node of the central section is extracted from the tidal dynamic field and used as the reference direction. If the integration threat value of the downstream fixed section is greater than that of the upstream fixed section, the reference direction is deflected 5 degrees downstream. If the integration threat value of the downstream fixed section is less than that of the upstream fixed section, the reference direction is deflected 5 degrees upstream. If the integration threat value of the downstream fixed section is equal to that of the upstream fixed section, the reference direction is not deflected. The corrected reference direction is then obtained and used as the transmission direction angle.

[0086] Multiply the power flow velocity value of the central grid node in the central segment by the integration threat value of that segment, and then multiply by the relative rate of change of the integration threat value over the past 15 minutes to obtain the original intensity.

[0087] The original intensity is divided by the product of the maximum integrated threat value and the maximum tidal current velocity of all sections of the entire channel at this moment, and the result is normalized to the 0-1 interval to obtain the transmission intensity. The transmission direction angle is combined with the transmission intensity to form the oil film threat transmission vector. The oil film threat transmission vector is used to predict the drift direction and intensity of the oil film to avoid traffic chaos in the downstream main channel due to sudden oil film drift.

[0088] In one embodiment, the oil film center point in the oil spill data is analyzed in conjunction with the tidal dynamic field to generate a disturbed motion vector of the oil film, and the method further includes:

[0089] Identify key route nodes in the target channel, analyze the impact of the oil slick threat transmission vector on key route nodes, and generate risk focus of key route nodes. Specifically, this includes: for the main channel axis of the target channel navigation area, calculate its rate of change of direction angle between continuous grid nodes, and mark the position with a rate of change of more than 30 degrees per hour as a turning point, otherwise not marking it; for the boundary of the navigation area, mark the point where two or more main channel axes intersect in the navigation area as the intersection point, and use the turning point and the intersection point as key route nodes.

[0090] Calculate the shortest path length from each key route node along the main channel axis to the center point of the oil slick, and use it as the channel approach distance; calculate the absolute difference between the transmission direction angle of the oil slick threat transmission vector and the direction angle of the line connecting the center point of the oil slick to the key route node, and obtain the angle deviation value; among them, the intersection point is given a basic sensitivity weight of 0.8 due to the complexity of traffic flow, and the turning point is given a basic sensitivity weight of 0.5 due to the limited space for ship maneuvering and the unstable course;

[0091] Multiply the reciprocal of the approach distance, 1 minus the angle deviation, the transmission strength of the oil film threat transmission vector, and the basic sensitivity weight, and normalize the product to the 0-1 range to obtain the risk focus degree of the key route node. The risk focus degree is used to identify the risk level of key nodes with complex traffic flow, so that traffic control can prioritize the smooth flow and safety of these key nodes.

[0092] The oil film threat transmission vector and risk focus are fused to generate the oil film disturbance motion vector. Specifically, this includes: identifying the critical route node with the highest risk focus from all critical route nodes and taking it as the dominant risk node; calculating the direction angle from the center point of the oil film to the dominant risk node and taking it as the node gravity direction angle.

[0093] Calculate the absolute value of the difference between the transmission direction angle of the oil film threat transmission vector and the nodal gravity direction angle. If the difference is greater than 180 degrees, subtract the value from 360 degrees to obtain the direction divergence angle. Then, multiply the risk focus value of the dominant risk node by the direction divergence angle and then by a coefficient of 0.005 to obtain the direction correction angle.

[0094] If the direction angle of the nodal gravitational force is clockwise from the direction angle of transmission, then the direction angle of transmission is added to this direction correction angle; otherwise, the direction correction angle is subtracted to complete the direction correction and obtain the direction angle of the disturbed motion.

[0095] The intensity of the disturbance motion is obtained by adding the transmission strength of the oil film threat transmission vector to the risk focus of the dominant risk node; the disturbance motion direction angle and the disturbance motion intensity are combined to generate the oil film disturbance motion vector representing the overall disturbance state of the oil film.

[0096] By collaboratively analyzing oil spill data and tidal dynamic fields, a disturbed motion vector of the oil slick is generated, accurately predicting the drift direction and intensity of the oil slick under tidal action. It determines the proportion of the oil slick encroaching on the main channel by calculating the channel occupancy rate, and reflects the degree of navigation obstruction by combining the navigation resistance potential energy value. At the same time, it identifies high-risk grids based on dynamic hazard coupling factors and locks the risk level of key route nodes through risk focus. This effectively solves the problem that existing static hazard zones cannot adapt to tidal effects and are difficult to predict oil slick changes, improves the accuracy of traffic control in narrow channels of ports with strong tides, ensures the smooth flow of key route nodes, and strengthens the effectiveness of maritime traffic control.

[0097] In one embodiment, the oil film center point is used as the tracking starting point, and the drift path of the oil film is analyzed based on the disturbed motion vector of the oil film and the tidal dynamic field to generate a predicted dangerous spatiotemporal set, including:

[0098] Based on the disturbed motion vector of the oil film, in the tidal dynamic field, multi-directional dynamic analysis is performed starting from the center point of the oil film to generate a path cell sequence. Specifically, it includes: generating the direction of three propagation branches based on the disturbed motion direction angle of the oil film disturbed motion vector. The three propagation branch directions are the disturbed motion direction angle, the disturbed motion direction angle plus 15 degrees, and the disturbed motion direction angle minus 15 degrees, respectively.

[0099] With a time step of 15 minutes, starting from the grid node where the oil film center point is located, the tracking is performed step by step for each propagation branch direction. Within each time step, the cosine value of the angle between the current flow direction angle and the current tracking direction of each of the eight neighboring grid nodes is calculated. The cosine value of the angle is multiplied by the traffic impact index of the neighboring grid node to obtain the step preference value. The grid node with the largest step preference value is selected as the target grid node for the next movement.

[0100] The current grid node, the target grid node, the current time step, and the step preference value are recorded together as a path cell. This process is repeated for each propagation branch, advancing eight time steps in succession. For each propagation branch, a sequence of eight path cells arranged in chronological order is generated, which is called the path cell sequence. The path cell sequence is used to simulate the potential drift path of the oil film in different directions, covering all the ranges where the oil film may spread, and avoiding missing potential dangerous areas.

[0101] For each path cell in the path cell sequence, the tidal dynamic field and the navigation resistance potential energy value are integrated to generate the drift stress intensity representing the deformation and migration of the oil film. Specifically, for each path cell in the path cell sequence, for the current grid node and the target grid node of the path cell, the velocity and flow direction angle in the tidal dynamic field corresponding to the current time step are decomposed into eastward and northward components, respectively. The difference between the eastward and northward components of the two nodes are calculated. The square root of the sum of the squares of these two differences is used to obtain the flow field deformation gradient. The flow field deformation gradient reflects the tensile and shear stress borne by the oil film due to the non-uniformity of the flow field.

[0102] Obtain the navigation resistance potential energy value of the fixed segment associated with the current grid node and the target grid node at the current time, and calculate the absolute value of the difference between the two navigation resistance potential energies to obtain the spatial change rate of resistance potential energy.

[0103] Multiply the flow field deformation gradient by the spatial rate of change of the impeding potential energy to obtain the original stress value; multiply the mean flow velocity of all grid nodes in the entire target channel at this moment by the standard deviation of the impeding potential energy values ​​of all fixed sections to obtain the potential flow value; divide the original stress value by the potential flow value and normalize the result to the 0-1 interval to generate the drift stress intensity representing the deformation and migration of the oil film at this node.

[0104] Based on the path cell sequence and drift stress intensity, the state of the oil film during its motion is analyzed, and the spatiotemporal diffusion entropy of the oil film is generated. Specifically, this includes: extracting the drift stress intensity sequence of the eight path cells of each propagation branch, calculating the coefficient of variation of the drift stress intensity sequence to represent the temporal fluctuation within the propagation branch; simultaneously, at the same time step, calculating the standard deviation of the drift stress intensity among the three propagation branches to represent the spatial heterogeneity among the propagation branches; and adding the coefficient of variation of each propagation branch to the average standard deviation of all propagation branches to obtain the disorder index of that propagation branch.

[0105] The entropy values ​​of the disorder indices of the three propagation branches are calculated as follows: divide the disorder index of each propagation branch by the sum of the three, take the natural logarithm and then the opposite of the logarithm, and then multiply by the quotient to obtain the calculation result. Add the three calculation results to obtain the oil film spatiotemporal diffusion entropy. The oil film spatiotemporal diffusion entropy reflects the uncertainty of the state of the oil film when it diffuses along different paths. The higher the value, the more chaotic the diffusion and the more difficult the prediction.

[0106] In one embodiment, taking the center point of the oil film as the tracking starting point, and based on the disturbed motion vector of the oil film and the tidal dynamic field, the drift path of the oil film is analyzed to generate a predicted dangerous spatiotemporal set, which further includes:

[0107] Based on the path cell sequence, the drift stress intensity and oil film spatiotemporal diffusion entropy are analyzed to generate a dynamic risk load. Specifically, for the eight path cells of each propagation branch, starting from the first path cell, its drift stress intensity is used as the cumulative risk value. For subsequent path cells, their cumulative risk values ​​are recalculated: the cumulative risk value of the previous path cell is multiplied by a decay coefficient of 0.85, and then the drift stress intensity of the current path cell is added to obtain the cumulative risk value of the path cell.

[0108] Calculate the cosine of the angle between the direction from the grid node of the previous path cell to the current grid node of the current path cell and the power flow direction angle of the current grid node to obtain the modulation factor of each path cell. If the cosine of the angle is less than 0, the modulation factor is 0.

[0109] Multiply the cumulative risk value of the current path cell by its corresponding modulation factor to obtain the dynamic cumulative risk value of the path cell.

[0110] The dynamic cumulative risk values ​​of the eight path cells of the propagation branch are added together to obtain the total cumulative risk of the branch. The total cumulative risk of the branch is multiplied by the spatiotemporal diffusion entropy of the oil film and the product is normalized to the 0-1 interval to generate the dynamic risk load of the propagation branch. The dynamic risk load is used to distinguish the risk level of different drift paths.

[0111] The dynamic risk load is analyzed to extract regions where the risk value continuously exceeds a preset threshold and has spatiotemporal continuity. A predicted hazard spatiotemporal set containing dangerous periods, dangerous regions, and dangerous levels is generated. Specifically, this includes: using the average dynamic risk load of all propagation branches as the preset threshold; for each propagation branch's path cell sequence, checking the grid nodes corresponding to each path cell in turn; if the dynamic cumulative risk value of the path cell is greater than the preset threshold, then the grid node is marked as a risk node and its time step is recorded; otherwise, it is not marked.

[0112] Spatially, risk nodes of adjacent grid nodes within each time step are aggregated into a spatial cluster. Temporally, in adjacent time steps (15 minutes apart), spatial clusters with a center distance of less than 150 meters are connected to form spatiotemporally continuous risk clusters. Each risk cluster must last for at least two consecutive time steps.

[0113] Then, the maximum value of the dynamic risk load of all propagation branches within the risk cluster is analyzed. If the maximum value is < 0.75, it is a level 1 hazard; if 0.75 ≤ maximum value ≤ 0.9, it is a level 2 hazard; if the maximum value is > 0.9, it is a level 3 hazard. The severity is level 3 hazard > level 2 hazard > level 1 hazard, thus obtaining the hazard level of each risk cluster.

[0114] Finally, the time range from the start to the end covered by each risk cluster is taken as the dangerous period, and the range formed by the coordinates of all nodes in the risk cluster is taken as the dangerous area. Combining the dangerous period, dangerous area and dangerous level generates the predicted dangerous spatiotemporal set.

[0115] By using the center point of the oil film as the starting point for tracking, and combining the disturbed motion vector of the oil film with the tidal dynamic field to generate a multi-branch path cell sequence, and accurately reflecting the deformation and migration state and diffusion uncertainty of the oil film through the drift stress intensity and the spatiotemporal diffusion entropy of the oil film, a predicted hazardous spatiotemporal set containing hazardous periods, hazardous areas, and hazardous levels is generated. This effectively covers the potential drift range of the oil film, solves the problem that existing static hazardous areas cannot predict oil film changes under the influence of tides, improves the effectiveness of traffic control in the event of an oil spill in a narrow channel of a port with strong tides, and ensures navigation order and safety.

[0116] In one embodiment, based on the predicted hazard spatiotemporal set, a dynamic predicted hazard zone is drawn, and traffic control instructions are issued, including:

[0117] The analysis of the predicted hazardous spatiotemporal set yields the hazardous area weight coefficients. Specifically, this includes: using the current time as a benchmark, calculating the time difference between the start time of the hazardous period in the predicted hazardous spatiotemporal set and the current time, reciprocating it and normalizing it to the 0-1 interval to obtain the time urgency factor, where the time difference is in hours.

[0118] Calculate the average value of the difference in dangerous area between adjacent time steps within the dangerous period of the risk cluster, and normalize the absolute value of the average value to the 0-1 interval to obtain the spatial expansion factor; calculate the standard deviation of the coordinates of all grid nodes within the risk cluster, and normalize the reciprocal of the standard deviation to the 0-1 interval to obtain the structural compactness factor.

[0119] Specifically, for the hazard levels, the basic risk value is 0.5 for Level 1 hazard, 0.7 for Level 2 hazard, and 0.9 for Level 3 hazard.

[0120] Multiply the basic risk value, time urgency factor, spatial expansion factor, and structural tightness factor together, and normalize the result to the 0-1 range to obtain the hazard zone weight coefficient of the risk cluster. The hazard zone weight coefficient is mainly used to reflect the control priority of different hazard zones.

[0121] The hazard zone weight coefficient is matched with the tidal dynamic field to generate a dynamic boundary correction factor. Specifically, for each risk cluster, for the center point of its hazard zone, the traffic impact index and flow direction angle of the grid node corresponding to the center point are extracted from the tidal dynamic field; the farthest points in eight directions are evenly selected as edge points at the edge of the hazard zone, and the traffic impact index and tidal flow velocity of these edge points at the current time are obtained respectively.

[0122] The average difference between the traffic impact index at the center point and the traffic impact indices at the eight edge points is calculated to obtain the center-edge difference value; the standard deviation of the tidal flow velocity at the eight edge points is calculated to obtain the edge velocity dispersion; and the average difference between the flow direction angle at the center point and the direction angle of the line connecting each edge point to the center point is calculated to obtain the flow direction divergence.

[0123] Multiplying the center-edge difference value, edge velocity dispersion, and flow direction divergence together yields the flow field deformation factor, which reflects the potential deformation effect of the flow field on the boundary of the hazardous area.

[0124] Finally, the basic risk value corresponding to the hazard level of the risk cluster is multiplied by the flow field deformation factor, and the square root of the product is normalized to the 0-1 interval to obtain the dynamic boundary correction factor of the risk cluster. The dynamic boundary correction factor is used to dynamically adjust the boundary of the danger zone in combination with the tidal flow characteristics to avoid the deviation between the boundary of the danger zone and the actual oil film position caused by the tidal flow change, and to ensure the accuracy of the control area.

[0125] In one embodiment, the process of drawing a dynamic predicted hazard zone based on the predicted hazard spatiotemporal set and issuing traffic control instructions further includes:

[0126] Based on the dynamic boundary correction factor and the hazard zone weight coefficient, the dynamic predicted hazard zone that evolves over time is plotted. Specifically, for each risk cluster in the predicted hazard spatiotemporal set, the outline of the hazard zone of that risk cluster at the current moment is used as the initial outline.

[0127] Based on the dangerous period of the risk cluster, it is divided into several 15-minute prediction sub-periods. For each prediction sub-period, the difference between the midpoint of the prediction sub-period and the starting time of the risk cluster is divided by the total length of the dangerous period to obtain the period evolution coefficient of the prediction sub-period. The difference and the total length are both in hours.

[0128] For each forecast sub-period, the geometric mean of the basic risk value of the risk level of its risk cluster and the dynamic boundary correction factor is multiplied by the time period evolution coefficient of that forecast sub-period to obtain the dynamic expansion radius of the risk area under each forecast sub-period.

[0129] For each boundary point on the initial profile, the tidal flow direction angle of the risk cluster center point in the current prediction sub-period is converted into the tidal flow dominant direction vector. The tidal flow dominant direction vector only indicates direction and has a length of 1. The direction directly pointing from the risk cluster center point to the current boundary point is calculated and converted into the initial radiation direction vector. The risk cluster center point is the center point of the danger zone.

[0130] The weight of the dominant tidal current direction vector is set to 0.6, and the weight of the initial radiation direction vector is set to 0.4. The components of the dominant tidal current direction vector and the initial radiation direction vector are multiplied by their respective weights and then added together to obtain the composite direction vector.

[0131] Among them, the dominant tidal direction vector has a greater impact on the final direction, so its weight is 0.6, while the initial radiation direction vector is mainly used to maintain the similarity between the shape of the danger zone and the initial contour and to prevent excessive deformation, so its weight is 0.4.

[0132] Each boundary point is moved by the distance of the dynamic expansion radius along the direction of the synthetic direction vector to obtain a new boundary point for the predicted sub-period. Connecting all the new boundary points constitutes the dynamic hazard zone outline for the sub-period.

[0133] Based on the hazard level of the risk cluster, the outline of the dynamic hazard zone for each predicted sub-period is filled with warning bands of different colors and transparency. Level 1, Level 2, and Level 3 hazards correspond to semi-transparent yellow, orange, and red, respectively. The outlines of all predicted sub-periods are superimposed in chronological order to form a dynamic predicted hazard zone that evolves continuously over time. The dynamic predicted hazard zone is used to show the drift trajectory and hazard range of the oil slick in the next 24 hours, which helps ships plan detour routes in advance, thereby achieving the effect of maritime traffic control.

[0134] Based on the dynamic prediction of danger zones, traffic control instructions are issued, specifically including: for dynamically predicted danger zones, traffic control instructions are issued, which prohibit approaching within 500 meters of the dynamically predicted danger zone.

[0135] By analyzing and predicting the spatiotemporal set of hazards, the weight coefficient of the hazard zone is calculated, accurately distinguishing the priority of control for different hazard zones. Combined with the tidal dynamic field, a dynamic boundary correction factor is generated to realize the dynamic adjustment of the hazard zone boundary with the tidal characteristics, avoiding the deviation between the boundary and the actual oil film position. Based on this, the dynamically predicted hazard zone can display the oil film drift trajectory and hazard range for the next 24 hours according to the time evolution. The risk level is intuitively presented through graded warning color bands, and a traffic control instruction prohibiting approach within 500 meters is issued. This solves the problem that static hazard zones cannot adapt to the tidal influence, thereby guiding ships to plan detour routes in advance, preventing traffic chaos in the main channel, improving the pertinence of traffic control in narrow channels of ports with strong tides, and ensuring navigation order and safety.

[0136] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An intelligent sensing system for unmanned vessel oil spill accidents, characterized in that, include: The tidal current analysis unit is used to acquire gridded tidal current forecast data of the target channel in real time from the maritime data center for at least two tidal cycles in the future, analyze the tidal current forecast data to obtain the tidal dynamic field, and the target channel is the port channel. The oil film analysis unit is used to acquire oil spill data sensed by the unmanned vessel in real time, and to perform joint analysis of the oil film center point in the oil spill data with the tidal dynamic field to generate the disturbed motion vector of the oil film, including: Obtain the navigation area of ​​the target channel; Based on the center point of the oil film, the overlapping area between the current contour boundary of the oil film and the navigation area is analyzed to generate the channel occupancy rate; The navigation resistance potential energy value is obtained by calculating the channel occupancy rate and oil film thickness; The traffic analysis unit, using the center point of the oil film as the starting point for tracking, and based on the disturbed motion vector of the oil film and the tidal dynamic field, analyzes the drift path of the oil film and generates a predicted spatiotemporal set of hazards, including: Based on the disturbed motion vector of the oil film, a multi-directional dynamic analysis is performed starting from the center point of the oil film in the tidal dynamic field to generate a path cell sequence. For each path cell in the path cell sequence, the tidal dynamic field and the navigation resistance potential energy value are fused to generate the drift stress intensity representing oil film deformation and migration. Based on the path cell sequence and drift stress intensity, the state of the oil film during the motion process is analyzed, and the spatiotemporal diffusion entropy of the oil film is generated. Traffic control units are used to map dynamic predicted hazard zones based on predicted spatiotemporal sets of hazards and issue traffic control commands.

2. The intelligent sensing system for unmanned vessel oil spill accidents according to claim 1, characterized in that, Analyzing tidal current forecast data yields the tidal dynamic field, including: Analyze the tidal current forecast data and calculate the accessibility of each grid node at different time points; The intensity of tidal current hazard is obtained by analyzing the abrupt changes in flow direction angle and velocity difference between adjacent grid nodes; By analyzing the vortex intensity of each grid node and combining the vortex intensity with the flow velocity from the tidal current forecast data, vortex-induced hazard energy that poses a potential threat to ship stability is generated.

3. The intelligent sensing system for unmanned vessel oil spill accidents according to claim 2, characterized in that, Analyzing tidal current forecast data yields the tidal dynamic field, which also includes: By integrating accessibility, tidal current hazard intensity, and vortex hazard energy, a comprehensive traffic impact value is generated for each grid node. Based on the comprehensive traffic impact value, the degree of traffic impact is assessed, and a tidal dynamic field is generated.

4. The intelligent sensing system for unmanned vessel oil spill accidents according to claim 3, characterized in that, By co-analyzing the oil film center point in the oil spill data with the tidal dynamic field, a disturbed motion vector of the oil film is generated, including: Analyze the navigation resistance potential energy and tidal dynamic field, and calculate the dynamic hazard coupling factor of each grid node; The coupling factor between the navigation resistance potential energy value and the dynamic hazard is calculated, and the transmission direction and intensity of the oil film along the target channel under the action of tidal current are analyzed to generate the oil film threat transmission vector.

5. The intelligent sensing system for an unmanned vessel oil spill accident according to claim 4, characterized in that, The oil spill data's oil film center point is analyzed in conjunction with the tidal dynamic field to generate the oil film's disturbed motion vector. This also includes: Identify key route nodes in the target channel, analyze the impact of oil slick threat transmission vectors on key route nodes, and generate risk focus of key route nodes; The oil film threat transmission vector and risk focus are fused to generate the oil film disturbance motion vector.

6. The intelligent sensing system for an unmanned vessel oil spill accident according to claim 1, characterized in that, Using the center point of the oil film as the starting point for tracking, and based on the disturbed motion vector of the oil film and the tidal dynamic field, the drift path of the oil film is analyzed to generate a predicted hazardous spatiotemporal set, which also includes: Based on the path cell sequence, the drift stress intensity and oil film spatiotemporal diffusion entropy are analyzed to generate dynamic risk loads; The dynamic risk load is analyzed to extract areas where the risk value continuously exceeds the preset critical value and has spatiotemporal continuity, generating a predicted spatiotemporal set of hazards that includes dangerous periods, dangerous areas and dangerous levels.

7. The intelligent sensing system for unmanned vessel oil spill accidents according to claim 6, characterized in that, Based on the predicted hazard spatiotemporal set, a dynamic predicted hazard zone is drawn, and traffic control instructions are issued, including: By analyzing the predicted hazard spatiotemporal set, the hazard zone weight coefficients are obtained; By matching the weight coefficient of the danger zone with the tidal dynamic field, a dynamic boundary correction factor is generated.

8. The intelligent sensing system for unmanned vessel oil spill accidents according to claim 7, characterized in that, Based on the predicted spatiotemporal set of hazards, a dynamic predicted hazard zone is drawn, and traffic control instructions are issued. This also includes: Based on the dynamic boundary correction factor and the hazard zone weight coefficient, the dynamic predicted hazard zone is plotted to evolve over time. Traffic control instructions are issued based on dynamic predictions of danger zones.

Images