A smart waterway monitoring and management system based on light poles

Abstract

This invention belongs to the field of intelligent maritime surveillance technology, specifically a smart waterway monitoring and management system based on light beacons. It synchronously acquires multi-source heterogeneous information about vessels through integrated light beacon equipment, performs spatiotemporal registration and target association processing on this information, generates dynamic vessel trajectories, and extracts static attributes, motion behavior characteristics, visual activity characteristics, and environmental interaction characteristics. Utilizing a multi-evidence chain collaborative decision-making mechanism, it accurately identifies four fundamentally different vessel states: normal anchoring, illegal anchoring, active navigation, and uncontrolled dragging of anchor. This achieves a leap from state perception to behavioral cognition, significantly improving the accuracy of vessel anchoring identification and its robustness in complex scenarios. Based on the identification results, it triggers differentiated alarm strategies matching the state, generating handling instructions with reasons for the state determination, thereby greatly optimizing the allocation of surveillance resources and improving emergency response efficiency.

Description

Technical Field

[0001] This invention belongs to the field of intelligent waterway monitoring technology, specifically a waterway intelligent checkpoint monitoring and management system based on light poles. Background Technology

[0002] The booming development of the water transport industry has created an urgent need for more refined and intelligent maritime supervision. Traditional light beacons, as navigation aids widely deployed in key waterways, have natural advantages such as reliable power supply and fixed geographical locations. In recent years, they have been explored as sensing nodes to build a new type of supervision model known as intelligent waterway checkpoints.

[0003] Existing methods for monitoring vessels in waterways based on light beacons mostly involve integrating Automatic Identification System (AIS) receivers, radar sensors, and video surveillance equipment onto the light beacons. AIS acquires vessel identity and dynamic information, radar compensates for target detection when AIS signals are missing, and video is used for manual confirmation.

[0004] However, this existing method has significant limitations when dealing with scenarios requiring deep behavior recognition, such as illegal anchoring, specifically: 1. Existing technologies lack the ability to distinguish the nature of ship anchoring behavior. They can only make a preliminary identification of whether a ship is stationary or at low speed, but cannot further distinguish the specific nature of the behavior corresponding to that state. For example, they cannot effectively distinguish the essential differences between legal anchoring, illegal anchoring, dragging anchor due to loss of control and normal navigation, resulting in insufficient targeting and intelligence of supervision.

[0005] 2. Existing anchoring identification mechanisms have low accuracy and reliability, relying on a single data source and simple threshold rules for judgment. For example, if the AIS speed is zero or the radar target is stationary, it is easy to make misjudgments or omissions due to environmental interference or deliberate ship manipulation. It lacks the ability to comprehensively analyze the ship's physical characteristics, environmental coupling effects, and multi-dimensional behavioral evidence, resulting in low regulatory efficiency and poor alarm credibility. Summary of the Invention

[0006] To overcome the shortcomings of the prior art, this invention provides a smart waterway monitoring and management system based on light stakes, which can effectively solve the problems mentioned in the prior art.

[0007] The objective of this invention can be achieved through the following technical solution: a smart waterway monitoring and management system based on light stakes, comprising: a multi-source data sensing module, a behavior feature deconstruction module, a ship status identification module, and a disposal module.

[0008] The multi-source data perception module is connected to the behavior feature deconstruction module, the behavior feature deconstruction module is connected to the ship status identification module, and the ship status identification module is connected to the disposal module.

[0009] The multi-source data sensing module, based on the integrated lighting equipment, synchronously acquires the ship's AIS information, radar trajectory information, visual image information, and real-time wind flow information.

[0010] The behavior feature deconstruction module performs spatiotemporal registration and target association processing on the information to generate ship dynamic trajectory and extract composite feature sequences. The sequences include at least ship static attributes, motion behavior features, visual activity features and environmental interaction features.

[0011] The ship status identification module, based on the composite feature sequence, sequentially performs main and auxiliary evidence chain analysis and status determination, and outputs ship status identification results including normal anchoring, illegal anchoring, active navigation, and uncontrolled dragging of anchor through multi-evidence chain collaborative decision-making.

[0012] The processing module, based on the identification result, triggers a differentiated alarm strategy that matches the state and generates a processing instruction with a reason for the state determination.

[0013] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects: (1) By constructing a multi-evidence chain collaborative decision-making mechanism, this invention can accurately distinguish four essential behavioral states: normal anchoring, illegal anchoring, active navigation, and uncontrolled dragging of anchor. It can accurately identify the nature of ship anchoring behavior, and is no longer a binary judgment of anchoring or non-anchoring. Instead, it can understand the intention and nature behind the behavior, provide a fundamental basis for differentiated handling, and realize the leap from state perception to behavior cognition.

[0014] (2) This invention breaks through the limitations of existing technologies that rely on a single data source and simple thresholds. By systematically integrating ship static attributes, motion behavior characteristics, visual activity characteristics and environmental interaction characteristics, it effectively overcomes misjudgments and omissions caused by environmental interference or malicious deception, thereby significantly improving the accuracy of ship anchoring identification and its robustness in complex scenarios.

[0015] (3) This invention accurately maps different ship status identification results with disposal strategies, constructs a differentiated disposal closed loop based on status identification, greatly optimizes the allocation of regulatory resources and improves emergency response efficiency. Attached Figure Description

[0016] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the module connection of the present invention.

[0018] Figure 2 This is a logical diagram illustrating the multi-evidence chain collaborative decision-making process of the present invention.

[0019] Figure 3 This is a schematic diagram illustrating the inclusion detection of the current position of a ship and the polygonal area of ​​an electronic fence using the ray method of the present invention. Detailed Implementation

[0020] 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.

[0021] Reference Figure 1 As shown, the present invention provides a waterway intelligent checkpoint monitoring and management system based on light stakes, including: a multi-source data sensing module, a behavior feature deconstruction module, a ship status identification module, and a disposal module.

[0022] The multi-source data perception module is connected to the behavior feature deconstruction module, the behavior feature deconstruction module is connected to the ship status identification module, and the ship status identification module is connected to the disposal module.

[0023] The multi-source data sensing module, based on the integrated light beacon equipment, simultaneously acquires the ship's AIS information, radar trajectory information, visual image information, and real-time wind flow information.

[0024] Although the preceding modules have gathered heterogeneous information from multiple sources, direct fusion may lead to trajectory breakage or feature misjudgment due to differences in acquisition frequency, system delay and clock reference among the information sources, inconsistent geographical coordinate reference systems of the data, and the possibility that the same ship target may use temporary, isolated or different identifiers in different information sources.

[0025] Therefore, the behavior feature deconstruction module first performs spatiotemporal registration and target association processing on the information, aiming to eliminate spatiotemporal bias and correctly associate observation points from different data sources with the same ship target, ensuring the accuracy and continuity of subsequent feature extraction. The specific implementation process is as follows: A linear interpolation extrapolation algorithm based on reference time is used to align the timestamps of each information source: First, the timestamps of all information are uniformly converted to UTC time. Then, taking one of the information sources as the main time axis, linear interpolation is performed between the two nearest main time axis observation points before and after each observation point of other information sources to calculate and calibrate the precise time position of the observation point on the main time axis, thereby achieving time synchronization.

[0026] By establishing a unified planar projection benchmark and converting the position information of all observation points to the planar rectangular coordinate system according to preset projection parameters, the conversion from WGS84 latitude and longitude to UTM coordinate system can be exemplified in specific implementation.

[0027] This generates a ship's dynamic trajectory and extracts a composite feature sequence, which includes at least the ship's static attributes, motion behavior features, visual activity features, and environmental interaction features.

[0028] Based on this, in a preferred embodiment of the present invention, the motion behavior features include the ground velocity stability state and the displacement distribution state, and the extraction process is as follows: A time window is set, and the distribution of ship position points within the window in the radar trajectory information is geometrically statistically analyzed to determine the geometric center of the ship position point cluster. The radius of the minimum enclosing circle containing a specific percentage of ship position points is calculated, and the displacement dispersion state is quantitatively characterized by the radius value.

[0029] Based on the radar trajectory information, statistical analysis is performed on the sample mean and standard deviation of the ship's ground speed within a continuous time window. The combination relationship between the mean and standard deviation after removing instantaneous fluctuation interference is used to characterize the stable state of the ship's ground speed.

[0030] The aforementioned statistical analysis of ship-to-ground speeds within a continuous time window requires a more refined execution process, including: First, a sliding window mean filter is applied to the original velocity sequence within the time window to smooth high-frequency random fluctuations.

[0031] Subsequently, median filtering is used to remove outliers from the smoothed velocity sequence.

[0032] Finally, based on the filtered speed data, the sample mean and sample standard deviation are calculated, and the combination of the two is used to characterize the ship's ground speed stability.

[0033] In a preferred embodiment of the present invention, the extraction process of the visual activity features is as follows: Based on significant changes in the ship's motion state, keyframe image sequences of the ship are extracted from visual image information.

[0034] To extract the aforementioned keyframe images, the keyframe extraction condition is that the ship's ground speed is lower than the ship's speed threshold and the duration exceeds a single set time window. The preset speed threshold can be specifically referred to as half of the ship's normal sailing speed, which is used to characterize the possibility of potential anchoring. Then, within the time period that meets the extraction condition, image frames are extracted from the video stream at fixed time intervals, and the extracted image frames are quality-screened to remove severely blurred or overexposed frames caused by weather, lighting, or other factors, thus forming the final keyframe image sequence.

[0035] Anchor chain target detection is applied to a preset area of ​​the ship deck in the keyframe image to identify the visibility and physical shape of the anchor chain and determine whether the deck anchor chain is in the extended or retracted state.

[0036] It should be added that the specific process for determining whether the deck anchor chain is in the extended or retracted state includes: searching relevant ship design manuals based on the ship type, determining the typical layout area of ​​the deck anchor chain equipment, and thus delineating the corresponding region of interest in the image.

[0037] An anchor chain target detection model trained on an anchor chain image dataset is used to scan and detect the region of interest.

[0038] Establish an image coordinate system with the center of the anchor chain hole as the origin, and detect the outline shape of the anchor chain and its relative positional relationship with the anchor chain hole.

[0039] When the main body of the anchor chain remains outside the threshold range around the anchor chain hole and the outline of the anchor chain exhibits a suspended characteristic extending towards the water surface, the anchor chain is determined to be in the released state.

[0040] When the main body of the anchor chain is concentrated within the threshold range around the anchor chain hole, or when only the outline of the anchor chain stored close to the deck is detected, it is determined that the anchor chain is in the retracted state.

[0041] The threshold range around the anchor chain hole is a circular area with the center of the anchor chain hole as its center and a radius adaptively determined based on the ship's own dimensions. Specifically, the radius is the product of the ship's total length and a preset proportional coefficient. The empirical value range of the preset proportional coefficient can be set as follows: .

[0042] The anchor chain's outline exhibits a hanging characteristic extending towards the water surface. Specifically, the quantification process is as follows: The skeleton of the detected anchor chain is extracted to obtain its centerline. The average tangent direction angle of this centerline near the starting segment of the anchor chain hole is calculated. When the angle between the average tangent direction angle and the horizontal plane satisfies the condition that the centerline is in the middle of the anchor chain hole, the centerline is considered to be in the middle of the anchor chain hole. When the trajectory is considered to conform to the suspended characteristics extending towards the water surface, it is considered to meet the requirements of the release trajectory.

[0043] The theoretical basis for setting the above-mentioned angle range is that when the anchor chain is released, the part near the anchor chain hole will be pulled down by the weight of the anchor chain and anchor below, forming a clear, downward sag curve at the beginning of the curve. If the angle is too small, such as close to 0°, it may indicate that the anchor chain has not been truly released or is still lying flat on the deck. If the angle is too large, such as close to 90°, it does not conform to the laws of physics or may be a visual error. The angle range effectively covers the typical physical form of a ship when the anchor chain begins to be released, at common drafts and berth depths.

[0044] Simultaneously, moving target analysis is performed on keyframe image sequences. By using background subtraction and optical flow estimation methods, the spatial distribution and duration characteristics of ship wakes on the water surface are extracted, and the wakes are distinguished as stable linear morphology or dissipating eddy morphology.

[0045] It should be added that the process of determining whether the above-mentioned wake is a stable linear shape or a dissipating vortex shape is as follows: the foreground motion region in the keyframe image sequence is extracted by background subtraction.

[0046] The motion vector of each pixel in the foreground motion region is calculated by applying an optical flow estimation algorithm. If the standard deviation of the direction angle of the motion vector of a pixel in a certain foreground motion region is less than 15 degrees, the region is determined to be the ship body with rigid motion characteristics. If it is greater than 15 degrees and the vector space distribution shows a divergent or vortex pattern, the region is determined to be a wake region.

[0047] Morphological analysis is performed on the segmented wake region. The aspect ratio of the minimum bounding rectangle of the wake region is calculated, and the ratio of the wake region area to its convex hull area is used as the contour regularity index. The number of frames of the wake region is counted and converted into the actual duration. A wake morphology discrimination function is established, which is a linear weighted fusion of aspect ratio, contour regularity index and actual duration. If the function output value exceeds the preset value, the wake is characterized as a stable linear morphology; otherwise, it is a dissipating vortex morphology. The preset value can be obtained by function mapping with reference to the following rules: when the aspect ratio is greater than 3:1, the contour regularity is greater than 0.8 and the duration exceeds 5 seconds, it is judged as a stable linear morphology.

[0048] The optical flow estimation algorithm and background difference method used in the above supplementary content are existing technologies and will not be elaborated here.

[0049] The output consists of visual activity features composed of the deck anchor chain status and the wake morphology.

[0050] In a preferred embodiment of the present invention, the environmental interaction feature extraction process is as follows: Based on the ship type and tonnage in the ship's static attributes, the ship's equivalent side projection area and inertia class are determined.

[0051] It should be noted that the determination of the equivalent side projected area and inertia class of the above-mentioned ships is actually based on the mapping relationship database between ship types and typical ship type parameters established before the application system was developed, and the corresponding ship type coefficients are called according to the input ship type.

[0052] Based on ship tonnage, length, beam, and draft data, and combined with ship type coefficients, the equivalent side projection area above and below the waterline is calculated using empirical formulas pre-set in the database.

[0053] Based on the ship's tonnage and inertial parameters, the ship's inertial class is divided into three levels: high, medium, and low. The high inertial class corresponds to ships with a tonnage greater than 100,000 tons, the medium inertial class corresponds to ships with a tonnage of 10,000 to 100,000 tons, and the low inertial class corresponds to ships with a tonnage less than 10,000 tons.

[0054] Based on the equivalent side projection area, inertia level, and real-time wind flow information, the upper limit of the expected rate of change of the angle between the ship's bow and the wind flow direction in an unpowered state is calculated as the theoretical dynamic response threshold.

[0055] Calculate the statistical variance of the rate of change of the angle between the ship's heading and the wind direction as actually observed by the ship within a set time window.

[0056] The difference coefficient between the statistical variance of the actual angle change rate and the theoretical dynamic response threshold is quantified and used as the environmental interaction feature.

[0057] It should be noted that the above difference coefficient can be exemplarily the result of the ratio calculation of the statistical variance of the actual angle change rate to the theoretical dynamic response threshold. When the difference coefficient is greater than 1, it indicates that the actual change in the ship's heading exceeds the theoretical expectation under no-power conditions and conforms to the irregular swaying characteristics under anchor conditions. When the difference coefficient is less than or equal to 1, it indicates that the actual change in the ship's heading is within the theoretical expectation range and conforms to the controlled motion characteristics under navigation conditions.

[0058] This invention overcomes the limitations of existing technologies that rely on a single data source and simple thresholds. By systematically integrating ship static attributes, motion behavior characteristics, visual activity characteristics, and environmental interaction characteristics, it effectively overcomes misjudgments and omissions caused by environmental interference or malicious deception, thereby significantly improving the accuracy of ship anchoring identification and its robustness in complex scenarios.

[0059] The ship status identification module, based on the composite feature sequence, sequentially performs main and auxiliary evidence chain analysis and status determination, and outputs ship status identification results including normal anchoring, illegal anchoring, active navigation, and uncontrolled dragging of anchor through multi-evidence chain collaborative decision-making.

[0060] Reference Figure 2As shown, in a preferred embodiment of the present invention, the multi-evidence chain collaborative decision-making includes the following implementation steps: S1. Based on the aforementioned motion behavior characteristics and environmental interaction characteristics, perform main evidence chain analysis to quantify the ship's initial anchoring probability.

[0061] S2. Combine the visual activity features to perform auxiliary evidence chain analysis, provide confidence corroboration for the primary anchoring probability, and generate a comprehensive anchoring probability.

[0062] S3. Compare the overall anchoring probability with a preset probability threshold. If the threshold is not reached, determine that the ship is in a sailing state and execute S4; otherwise, determine that the ship is in an anchoring state and execute S5.

[0063] S4. By combining the wind pressure difference with the ship's static properties, estimate the ship's theoretical drift trajectory without power. If it matches the ship's actual drift trajectory, output the active navigation state; otherwise, output the out-of-control anchor dragging state.

[0064] S5. Determine the ship's current position based on the preset electronic fence layout. If the ship's current position is within a legal anchorage, output a normal anchorage status. If it is within a prohibited anchorage area or restricted anchorage, output an illegal anchorage status.

[0065] In a preferred embodiment of the present invention, the main chain of evidence analysis includes: If the minimum enclosing circle radius is less than a preset proportion of the ship's length, then the ship is identified as conforming to the displacement circle distribution pattern under anchored conditions, and a first contribution value is generated.

[0066] Based on the combination of the mean and standard deviation representing the steady state of the ship's speed relative to the ground, it is determined whether the ship's speed relative to the ground conforms to the low-fluctuation pattern approaching zero under anchored conditions. If it does, a second contribution value is generated.

[0067] Based on the difference coefficient in the environmental interaction features, it is determined whether the angle between the ship's bow and the wind current conforms to the irregular swaying pattern under anchoring conditions. If it does, a third contribution value is generated.

[0068] The first contribution value, the second contribution value, and the third contribution value are summed to obtain the initial anchoring probability.

[0069] It should be noted that the first, second, and third contribution values ​​mentioned above were all assigned values ​​before system development. The specific values ​​are based on the importance weight of each piece of evidence for anchoring judgment: the first contribution value directly reflects the stability of the ship's position, with an assigned value range of 40-50, and has the highest importance; the second contribution value reflects the ship's motion state, with an assigned value range of 20-30; and the third contribution value reflects the interaction between the ship and the environment, with an assigned value range of 15-25. This invention requires that when all three pieces of evidence are satisfied simultaneously, the initial anchoring probability is assigned a value of 80-90, in order to reserve room for increasing the probability for subsequent supplementary evidence chains.

[0070] In a preferred embodiment of the present invention, the auxiliary evidence chain analysis includes: If the anchor chain on the deck is identified as being released and the wake pattern is a stable linear pattern, the first confidence factor is generated.

[0071] Extract the ship's ground speed change sequence from the radar trajectory information, identify whether the ship exhibits a behavior pattern of continuously decreasing speed before reaching the current position, and if so, generate a second confidence factor.

[0072] It should be noted that the identification of the above-mentioned pattern of continuously decreasing speed is based on the following quantitative criteria: Capture the sequence of changes in the ship's ground speed within a set time window before the ship reaches its current position.

[0073] Linear regression analysis was performed on the velocity change sequence to calculate its slope value.

[0074] When the slope is negative and the coefficient of determination is greater than 0.7, it is determined to be a continuously decreasing pattern.

[0075] At the same time, it is required that the end value of the velocity sequence be lower than half of the starting value to ensure that the velocity is substantially reduced.

[0076] Based on the first confidence factor and the second confidence factor, the primary anchoring probability is positively weighted and corrected to generate the comprehensive anchoring probability.

[0077] The assignment and weighting adjustment of the first confidence factor and the second confidence factor adopt the following scheme: The first confidence factor involves visual evidence, which has high directness but may be affected by observation conditions. When both the anchor chain release state and the stable linear wake are identified simultaneously, the first confidence factor takes the maximum value of 0.06. When only one of the anchor chain release state or the stable linear wake is identified, the first confidence factor takes the median value of 0.03.

[0078] The second confidence factor involves evidence of behavioral patterns, which has good objectivity but may be ambiguous, so it is given a moderate weight of 0.04.

[0079] The sum of the two confidence factors is capped at 0.1 to ensure that the secondary chain of evidence plays an appropriate calibrating role without unduly affecting the preliminary conclusions drawn from the primary chain of evidence.

[0080] The weighted correction formula is as follows: 1 is added to the confidence factor, and then multiplied by the primary anchoring probability.

[0081] In a preferred embodiment of the present invention, the process of estimating the ship's theoretical drift trajectory without power includes: determining the ship's windward area and the hull side area below the waterline based on the ship's main dimensions and tonnage in the ship's static attributes.

[0082] Based on the determined area parameters and real-time airflow information, the resultant center and magnitude of the wind-induced drift force and the flow-induced drift force are calculated respectively.

[0083] Based on the center and magnitude of the resultant force, the unpowered theoretical drift vector of the composite ship's center of mass is obtained.

[0084] Based on the theoretical drift vector, the ship's current position is integrated over time to generate the theoretical drift trajectory without power.

[0085] In a preferred embodiment of the present invention, the ship's theoretical drift trajectory without power must match the ship's actual drift trajectory to satisfy the conditions that the dominant direction is spatially consistent and the displacement amplitude is positively proportional.

[0086] It should be noted that the above-mentioned process for determining the consistency of the dominant directional space is as follows: the theoretical drift trajectory and the actual drift trajectory are represented as displacement vectors respectively, the cosine similarity of the two vectors is calculated, and when the similarity is greater than 0.9, it is determined that the directional space is consistent.

[0087] The process for determining the displacement amplitude ratio is as follows: calculate the relative error between the theoretical displacement vector magnitude and the actual displacement vector magnitude. When the relative error is less than 20%, it is determined that the displacement amplitude is positively proportional.

[0088] In a preferred embodiment of the present invention, determining the current position of the ship based on a preset electronic fence layout includes: Acquire electronic nautical chart data containing information on legal anchorages, prohibited anchorages, and restricted anchorage boundaries.

[0089] The electronic nautical chart data is analyzed to construct an electronic fence spatial layer based on a geographic coordinate system.

[0090] The coordinates of the ship's current position are overlaid and calculated in real time with the electronic fence spatial layer.

[0091] The ray casting method is used to perform inclusion detection between the ship's current position and the polygonal area of ​​the electronic fence, and the specific electronic fence area type in which the ship is located is output.

[0092] It should be noted that this can be referred to Figure 3 As shown, the above-mentioned ray method is implemented as follows: a ray is drawn from the current position of the ship in any direction.

[0093] Calculate the number of all intersections between the ray and the boundary of a certain electronic fence area.

[0094] If the number of intersection points is odd, the ship's location is determined to be inside the electronic fence area; if the number is even, it is determined to be outside the area.

[0095] This invention, through the construction of a multi-evidence chain collaborative decision-making mechanism, can accurately distinguish four fundamentally different behavioral states: normal anchoring, illegal anchoring, active navigation, and uncontrolled dragging of anchor. This enables precise identification of the nature of ship anchoring behavior, moving beyond a binary judgment of simply anchoring or not anchoring. Instead, it allows for insight into the intent and nature behind the behavior, providing a fundamental basis for differentiated handling and achieving a leap from state perception to behavioral cognition.

[0096] The processing module triggers a differentiated alarm strategy that matches the status based on the identification result, and generates a processing instruction with a reason for the status determination.

[0097] It should be added that the above-mentioned differentiated alarm strategies specifically include: When the identification result indicates illegal anchoring, a drive-away alarm is triggered and a disposal work order is dispatched to the law enforcement unit.

[0098] When the vessel loses control and drags anchor, a high-risk safety alarm is triggered, and a warning message is broadcast to nearby vessels and rescue units.

[0099] During active navigation, the system performs routine tracking and recording.

[0100] During normal anchoring, the system records and files information without triggering any active alarms.

[0101] The reasons for determining the status attached to the above-mentioned disposal instructions include, but are not limited to: displacement circle distribution diagrams used to corroborate the anchoring status, bow-current relationship diagrams used to corroborate the hydrodynamic pattern relationship, and key image snapshots used to corroborate visual activity characteristics.

[0102] This invention accurately maps different ship status identification results with handling strategies, constructing a differentiated handling closed loop based on status identification, which greatly optimizes the allocation of regulatory resources and improves emergency response efficiency.

[0103] The above description is merely an example and illustration of the structure of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the structure of the invention or exceed the scope defined by the present invention, they should all fall within the protection scope of the present invention.

Claims

1. A smart waterway monitoring and management system based on light poles, characterized in that, include: The multi-source data sensing module, based on the integrated light beacon equipment, synchronously acquires the ship's AIS information, radar trajectory information, visual image information, and real-time wind flow information; The behavior feature deconstruction module performs spatiotemporal registration and target association processing on the information to generate ship dynamic trajectory and extract composite feature sequence, the sequence including at least ship static attributes, motion behavior features, visual activity features and environmental interaction features; The ship status identification module, based on the composite feature sequence, sequentially performs main and auxiliary evidence chain analysis and status determination, and outputs ship status identification results including normal anchoring, illegal anchoring, active navigation, and uncontrolled dragging of anchor through multi-evidence chain collaborative decision-making; The processing module, based on the identification result, triggers a differentiated alarm strategy that matches the state and generates a processing instruction with a reason for the state determination.

2. The intelligent waterway monitoring and management system based on light poles according to claim 1, characterized in that, The motion behavior features include the steady state of ground velocity and the displacement distribution state, and the extraction process is as follows: A time window is set, and geometric statistics are performed on the distribution of ship position points within the window in the radar trajectory information to determine the geometric center of the ship position point cluster. The radius of the minimum enclosing circle containing a specific percentage of ship position points is obtained, and the displacement dispersion state is quantitatively characterized by the radius value. Based on the radar trajectory information, statistical analysis is performed on the sample mean and standard deviation of the ship's ground speed within a continuous time window. The combination relationship between the mean and standard deviation after removing instantaneous fluctuation interference is used to characterize the stable state of the ship's ground speed.

3. The intelligent waterway monitoring and management system based on light poles according to claim 2, characterized in that, The extraction process of the visual activity features is as follows: Based on significant changes in the ship's motion state, keyframe image sequences of the ship are extracted from visual image information; Anchor chain target detection is applied to a preset area of ​​the ship deck in the keyframe image to identify the visibility and physical shape of the anchor chain and determine whether the deck anchor chain is in the extended or retracted state. Simultaneously, moving target analysis is performed on keyframe image sequences. By using background subtraction and optical flow estimation methods, the spatial distribution and duration characteristics of ship wakes on the water surface are extracted, and the wakes are distinguished as stable linear morphology or dissipating eddy morphology. The output consists of visual activity features composed of the deck anchor chain status and the wake morphology.

4. The intelligent waterway monitoring and management system based on light poles according to claim 3, characterized in that, The environmental interaction feature extraction process is as follows: Based on the ship type and tonnage in the ship's static attributes, determine the ship's equivalent side projection area and inertia class. Based on the equivalent side projection area, inertia level, and real-time wind flow information, the upper limit of the expected rate of change of the angle between the ship's bow and the wind flow direction in an unpowered state is calculated as the theoretical dynamic response threshold. Calculate the statistical variance of the rate of change of the angle between the ship's heading and the wind direction as actually observed by the ship within a set time window; The difference coefficient between the statistical variance of the actual angle change rate and the theoretical dynamic response threshold is quantified and used as the environmental interaction feature.

5. The intelligent waterway monitoring and management system based on light poles according to claim 4, characterized in that, The multi-evidence chain collaborative decision-making includes the following implementation steps: S1. Based on the aforementioned motion behavior characteristics and environmental interaction characteristics, perform main evidence chain analysis to quantify the ship's initial anchoring probability; S2. Combine the visual activity features to perform auxiliary evidence chain analysis, provide confidence corroboration for the primary anchoring probability, and generate a comprehensive anchoring probability; S3. Compare the overall anchoring probability with a preset probability threshold. If the threshold is not reached, determine that the ship is in a sailing state and execute S4. Otherwise, determine that the ship is in an anchoring state and execute S5. S4. By combining the wind pressure difference with the ship's static properties, estimate the ship's theoretical drift trajectory without power. If it matches the ship's actual drift trajectory, output the active navigation state; otherwise, output the out-of-control anchor dragging state. S5. Determine the ship's current position based on the preset electronic fence layout. If the ship's current position is within a legal anchorage, output a normal anchorage status. If it is within a prohibited anchorage area or restricted anchorage, output an illegal anchorage status.

6. The intelligent waterway monitoring and management system based on light posts according to claim 5, characterized in that, The main chain of evidence analysis includes: If the minimum enclosing circle radius is less than a preset proportion of the ship's length, then the ship is identified as conforming to the displacement circle distribution pattern under anchored conditions, and a first contribution value is generated. Based on the combination of the mean and standard deviation representing the steady state of the ship's ground speed, it is determined whether the ship's ground speed conforms to the low fluctuation pattern that approaches zero under anchored conditions. If it does, a second contribution value is generated. Based on the difference coefficient in the environmental interaction features, it is determined whether the angle between the ship's bow and the wind current conforms to the irregular swaying pattern under anchoring conditions. If it does, a third contribution value is generated. The first contribution value, the second contribution value, and the third contribution value are summed to obtain the initial anchoring probability.

7. The intelligent waterway monitoring and management system based on light poles according to claim 5, characterized in that, The auxiliary evidence chain analysis includes: If the anchor chain on the deck is identified as being in the extended state and the wake pattern is a stable linear pattern, the first confidence factor is generated. Extract the ship's ground speed change sequence from the radar trajectory information, identify whether the ship exhibits a behavior pattern of continuously decreasing speed before reaching the current position, and if so, generate a second confidence factor. Based on the first confidence factor and the second confidence factor, the primary anchoring probability is positively weighted and corrected to generate the comprehensive anchoring probability.

8. The intelligent waterway monitoring and management system based on light poles according to claim 5, characterized in that, The process for estimating the theoretical drift trajectory of a ship without power includes: Based on the ship's main dimensions and tonnage in its static properties, determine the ship's windward area and the hull side area below the waterline. Based on the determined area parameters and real-time airflow information, the center and magnitude of the resultant force of wind-induced drift force and flow-induced drift force are calculated respectively. Based on the center and magnitude of the resultant force, the unpowered theoretical drift vector of the composite ship's center of mass; Based on the theoretical drift vector, the ship's current position is integrated over time to generate the theoretical drift trajectory without power.

9. A smart waterway monitoring and management system based on light poles according to claim 5, characterized in that, The theoretical drift trajectory of a ship without power must match the actual drift trajectory of the ship to meet the conditions that the dominant direction is spatially consistent and the displacement amplitude is positively proportional.

10. A smart waterway monitoring and management system based on light poles according to claim 5, characterized in that, The determination of the ship's current position based on the preset electronic fence layout includes: Acquire electronic nautical chart data containing information on legal anchorages, prohibited anchorage zones, and restricted anchorage boundaries; The electronic nautical chart data is analyzed to construct an electronic fence spatial layer based on a geographic coordinate system; The coordinates of the ship's current position are overlaid and calculated in real time with the electronic fence spatial layer; The ray casting method is used to perform inclusion detection between the ship's current position and the polygonal area of ​​the electronic fence, and the specific electronic fence area type in which the ship is located is output.

Images